{"as_of":"2026-11-08T15:48:41.078933+00:00","build_id":"20261108-af50cab8676ded11","catalog_count":108,"catalog_revision":"c85255a17e571b2b77e7b38186dbe7c159e2b0552cd31fc3cc716d13e356d55b","contributing_observation_count":1272,"details":{"m-016da3c7828e329e":{"id":"m-016da3c7828e329e","in_window":true,"name":"Claude Haiku 5.5","observed_records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e5de55a774fc4f793b69c43a1c3f15d2","model_id":"m-016da3c7828e329e","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1620.0,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"1620","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row2:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-cdb1d51fae04:table0:row2:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"分","value":1620.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"f01303851f0ac5f8f962f37a067f5763","model_id":"m-016da3c7828e329e","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Knowledge work AA-Briefcase v1.1; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1578.0,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"1578","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row4:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-cdb1d51fae04:table0:row4:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":1578.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","model_id":"m-016da3c7828e329e","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"be88a9ace82ed4a1faf1b6eaa2b8a0bf","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":39.2,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"39.2%","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row11:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-cdb1d51fae04:table0:row11:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":39.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e2ba5f9af59c0998db11f28273d765a7","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":36.36,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"36.36%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":36.36},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","model_id":"m-016da3c7828e329e","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5f74b7f1bc8e1a0b82693e82428d74e5","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1441.38,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1441.38},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e4b023a9d06a1d503364382dbde1e0d0","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1523.33,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1523.33},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"48aa8f14043b6e1909ed8e6624348d53","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1114.41,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1114.41},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"214bdff55d5212e9a2a202b98eb6608c","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1367.13,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1367.13},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f63037f107bff704865ed9ebe0884cd2","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.31833333333333336,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31833333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3b1351d90b13ee3b33ac3b4d1ba7f4b8","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"799c5277fa75792cef21addf4b5dc0e6","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30366666666666664,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30366666666666664},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26ba087bf077278fb7cba6b0a9f40fdb","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.31733333333333336,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31733333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f983b976c1deefd0e29c60ff457feca1","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.610757946210269,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.610757946210269},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7cbe0c3cf4fffa96ec46dbe785179897","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6084549572219426,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6084549572219426},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae5f539edc14bfc9444bfef59e484e2c","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.595021541407372,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595021541407372},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e5b23a2d1f50414e9a80ddc7b3871c33","model_id":"m-016da3c7828e329e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.59619140625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59619140625},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e00698d26b1e6dcfd5609e19e4c8a5f","model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3372685311767238,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372685311767238},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57a6a58e67f34fbc5d05a26b7a3df975","model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35953889773428194,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35953889773428194},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a9eb8c73db9c99734b0b234d31c0de7","model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22910812181110643,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22910812181110643},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf280741919aee8a7710c5364568c9c8","model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2858961481429316,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2858961481429316},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9816a0d69d409a0896cabd3848bb59bf","model_id":"m-016da3c7828e329e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-haiku-5-5-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8d3fcd9050aa14ff77e18588232eeac4","model_id":"m-016da3c7828e329e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.180952380952381,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-haiku-5-5-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.180952380952381},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b90c8a4d69a85af7d420d3de2d9bbead","model_id":"m-016da3c7828e329e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-haiku-5-5-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"736379d5812d86c74d58e127e686a3db","model_id":"m-016da3c7828e329e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.2,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-haiku-5-5:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"335dc3727d4df9469c9d8f21b8bc04b1","model_id":"m-016da3c7828e329e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0714285714285714,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-haiku-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0714285714285714},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"64347fc93e768a048a007c64cb49d99a","model_id":"m-016da3c7828e329e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.232000505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232000505050505},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"63fbe38a7c32f64f1e27e1bed43687f2","model_id":"m-016da3c7828e329e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.299334090909091,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.299334090909091},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2ed4363afc1d6428ff188ccff76b7374","model_id":"m-016da3c7828e329e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0803666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0803666666666667},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","model_id":"m-016da3c7828e329e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e0c5b7a87cec2ac5e963e7ec7fc16885","model_id":"m-016da3c7828e329e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.16091994949495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16091994949495},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8766e12275bc386a0390857661b2f789","model_id":"m-016da3c7828e329e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.237,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.237},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cf5c6313488c3e2f67952c7e21597018","model_id":"m-016da3c7828e329e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.288,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.288},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"af869be60689a08bcfcf8ff82e88046a","model_id":"m-016da3c7828e329e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.141,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.141},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","model_id":"m-016da3c7828e329e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"24d27558810e3a3619fa63ba748cf91b","model_id":"m-016da3c7828e329e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.168,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.168},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ba07583c580418d100f3f0fba893fda","model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e24f03c35528622b99e8c1c7569c8e2a","model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.182,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f9572b6eb3289dc28f8c685386a8905a","model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a2d888552d8101005385f5cc4f527aa","model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8fe11106baa595570ce9b33abce78d5","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.217171717171717,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.217171717171717},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bbddb0051195d204c7d82370a21c3fcd","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.292929292929293,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7c176301a9a5702522096aa15abca045","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d85decab03c1cd3287927ba38c66b8ab","model_id":"m-016da3c7828e329e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.151515151515152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.151515151515152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"254457afaca82e5fa1fa571696a00395","model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"05db37988c471adae1fde5ffcedd958f","model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.475,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.475},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2313a02f9c4a1b6338a9e7d4b648288d","model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.215,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.215},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10750687ebb9102c142b25f367215093","model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.305,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c5e39f2851c6a5f2c7abbbda97788a","model_id":"m-016da3c7828e329e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1424.29,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1424.29},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"88d7d4149e7cdeaf81278fed4ff8d267","model_id":"m-016da3c7828e329e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1521.54,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1521.54},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"88c3b51db416d76a6d585a07441f9e74","model_id":"m-016da3c7828e329e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1130.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1130.36},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","model_id":"m-016da3c7828e329e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e4d8c9223ca8ba1c42b945637d118ae5","model_id":"m-016da3c7828e329e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1282.5,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1282.5},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-07","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":2570275},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":350},"family":"FrontierMath","row_ids":["21e96013106dd82d62904afe0c4f768b"],"score":38.89},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["0e248a56228a0b9ac7b762f2f3b00fc9"],"score":57.89},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5000},"family":"HLE","row_ids":["de9339207cccf727d5f73de793318c18"],"score":79.37},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1300},"family":"CritPt","row_ids":["b0085122202e14e07e4a784b0954985f"],"score":76.47}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Diamond","peer_count":64,"percentile":79.37,"percentile_exact":{"denominator":63,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","peer_count":52,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435}],"score":63.15,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":143451,"numerator":9916043},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2200},"family":"GDPval-AA","row_ids":["ac3f66c4919122d1f31985187718ed07"],"score":81.48},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":224},"family":"Terminal-Bench","row_ids":["7706b3202bdd1e09b6c5e4786d4bb3e4"],"score":74.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"AutomationBench","row_ids":["9f3e53e10beed094c8835d4e672cc44f"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6800},"family":"AA-Briefcase","row_ids":["6fc39f46271fccbf5e9ec9022c726f0f"],"score":88.31},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":100},"family":"OSWorld","row_ids":["5b785287811a33de601a7060b80bab3e"],"score":33.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2000},"family":"AA-AnalystAgent","row_ids":["a7b164121b0a9a4769bf158d9212d770"],"score":86.96}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2.1","peer_count":82,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":78,"percentile":88.31,"percentile_exact":{"denominator":77,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.1","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","peer_count":24,"percentile":86.96,"percentile_exact":{"denominator":23,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375}],"score":69.12,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":91,"numerator":6350},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":600},"family":"FrontierCode","row_ids":["cbf1fa28fdc821e9836816e2dbc59c73"],"score":85.71},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":700},"family":"Vals Code Migration","row_ids":["c1b5961685ee27b7ccd1b99becf1d743"],"score":53.85}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.0","peer_count":50,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"1.1","peer_count":27,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36}],"score":69.78,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":4819875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2975},"family":"GDP.pdf","row_ids":["a183b88682c41fc78cc2ad5e5485bbb7"],"score":78.29},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":100},"family":"SimpleQA Verified","row_ids":["a7e4020697adbd00e97487d7c55696e0"],"score":5.88},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"MultiChallenge","row_ids":["5d9b045cf9ce01ea061077e1b2e84c1e"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3000},"family":"AA-Omniscience","row_ids":["30bc019f7dc3e7314870de193c23616d","b11d60a588bf725879871b88bc8f70cb"],"score":56.6}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897}],"score":35.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"Chartography","row_ids":["888dd585273ea6255f0bebeb11182076"],"score":50.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4}],"score":50.0,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Diamond","peer_count":64,"percentile":79.37,"percentile_exact":{"denominator":63,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","peer_count":52,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2.1","peer_count":82,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":78,"percentile":88.31,"percentile_exact":{"denominator":77,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.1","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","peer_count":24,"percentile":86.96,"percentile_exact":{"denominator":23,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.0","peer_count":50,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"1.1","peer_count":27,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4}],"result_updated_at":"2026-10-09","score":59.06,"score_exact":{"denominator":127698358788,"numerator":7542473801321}}],"release_date":"2026-10-07","release_status":"published","release_verified":true,"source_id":"anthropic/claude-haiku-5-5"},"m-0510680735d3c79a":{"id":"m-0510680735d3c79a","in_window":true,"name":"GPT-5.6 Sol","observed_records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"3a2ab6999fa584f73c8d3dab12d76e7c","model_id":"m-0510680735d3c79a","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"ac0f80ad7501f371bea4ff3e","raw_value":37.3,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench:Terminal-Bench 4.0","source_record_id":"gpt6-astra-official:table2:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":37.3},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","model_id":"m-0510680735d3c79a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"b48576f95bc7cd0aae9b966cadaca0c7","model_id":"m-0510680735d3c79a","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench Science 0.1; preserve page footnotes and task-specific settings","protocol_id":"6e315c8a4731951529a013b5","raw_value":22.4,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench-Science:Terminal-Bench Science 0.1","source_record_id":"gpt6-astra-official:table3:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":22.4},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","model_id":"m-0510680735d3c79a","original_version_label":"v2","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"2f88107526bcb2b1d26a068b34f3a3d3","model_id":"m-0510680735d3c79a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":72.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"40f8d3f23ce7ba4f99898294659f1d24","model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":83.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":83.0},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"5b5d809739c95968ceacfb8058855082","model_id":"m-0510680735d3c79a","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":89.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":89.0},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"9e593dba9c9742a44e63ab3d6501ad09","model_id":"m-0510680735d3c79a","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":83.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"bbf54703736529a05510c77294f21819","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":18.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"737049649a2647ef4d75f801d25ae0a1","model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"94.6%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":94.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GPT-5.6 Sol / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"85b0ca5b600459af0704f6882d2f67f7","model_id":"m-0510680735d3c79a","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.9,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GPT-5.6 Sol"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GPT-5.6 Sol","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","model_id":"m-0510680735d3c79a","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"cc26eea60ec9309fc3beabf4435c173d","model_id":"m-0510680735d3c79a","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Agentic scientific research Terminal-Bench-Science 0.1³; preserve page footnotes and task-specific settings","protocol_id":"1975ca8c09f87d3fc7597c12","raw_value":22.4,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"22.4%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row14:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1","source_record_id":"official-0dd3bed5c7dd:table0:row14:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":22.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"fa113390898a4649fa21eed1a4997857","model_id":"m-0510680735d3c79a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":73.0,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"73.0","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":73.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d2394a533761e9936e4a49c1dfe512c8","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":29.7,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"29.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":29.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6-Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"723e50c3bda3d52daa8d895fc55553b6","model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":83.0,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"83.0","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3d74a21fb4db168c492ba73576bd0b2e","model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT5.6 Sol (max). Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":94.1,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"94.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":94.1},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"GPT 5.6 Sol (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"975826e2b4130a8d077f072f1973ca40","model_id":"m-0510680735d3c79a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":73.0,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"73.0","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":73.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"GPT 5.6 Sol (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"ebef37315791ccc04717ed5ba0766942","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":29.7,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"29.7","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":29.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"GPT 5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"e818a4b1e85e949ea72540313ee99c3c","model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":72.7,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"72.7","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":72.7},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"GPT 5.6 Sol (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"593853e47efc92fb5fb50665ad18805a","model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT 5.6 Sol (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":94.1,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"94.1","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":94.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6ee2b87b3178b7a9ed8fa714c84647bf","model_id":"m-0510680735d3c79a","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":39.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"39.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":39.9},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"8fac1a5b5e9d5b95efd26212af5182b5","model_id":"m-0510680735d3c79a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":73.0,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"73.0","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":73.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"b1f2d23ce08087c377528f7267508376","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":45.8,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"45.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":45.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"a730c185c3223a52793bf1eef3941811","model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column GPT-5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":94.1,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"94.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":94.1},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GPT-5.6 Sol (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6817c08660542a8ba2cec888e8e593bf","model_id":"m-0510680735d3c79a","original_version_label":"1.1; explicit developer footnote","profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":73.0,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"73.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column3","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":73.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GPT-5.6 Sol (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"35d3b6c21baca0074691fdf7baa39656","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":29.7,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"29.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column3","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":29.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GPT-5.6 Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c21818a023d8c8e1946a57f37d71c67d","model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":83.0,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"83.0 / 84.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column3","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GPT-5.6 Sol (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"f9238d0464b5e8da4ef9b0d43a936fa9","model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT-5.6 Sol (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":94.1,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"94.1","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column3","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":94.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GPT-5.6 Sol (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c608d42d9d5288f84ff9a46cf0c9f453","model_id":"m-0510680735d3c79a","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GPT-5.6 Sol (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":83.6,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"83.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column3","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":83.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GPT-5.6 Sol (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"62c8c73bb0229cb8c08fb44bb8f568ca","model_id":"m-0510680735d3c79a","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GPT-5.6 Sol (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.9,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column3","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"fad33698f014ebb1ac938c69d7c0798b","model_id":"m-0510680735d3c79a","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic scientific research Terminal-Bench-Science 0.1 [1]; preserve page footnotes and task-specific settings","protocol_id":"88023c6718be09c275d82cfc","raw_value":22.4,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"22.4%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row2:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1 [1]","source_record_id":"anthropic-fable51:table0:row2:column4","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":22.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"5ba80265411deda6cb00777e57f78095","model_id":"m-0510680735d3c79a","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":37.3,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"37.3%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row4:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"anthropic-fable51:table0:row4:column4","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":37.3},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"da2362332ab0039d90ab52612d3c3148","model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":19.6,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"19.6%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column4","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":19.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"d009cc9a0555aee5404869255b905141","model_id":"m-0510680735d3c79a","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":94.7,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"94.7","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":94.7},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","model_id":"m-0510680735d3c79a","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"fadaedd3ee1f675326bbb9970170cf61","model_id":"m-0510680735d3c79a","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":39.9,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"39.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":39.9},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"b7c2bea8fb6f0a31dd7c2532db1fe168","model_id":"m-0510680735d3c79a","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.9,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"f34f41dbda882df5d5652b5a2253e938","model_id":"m-0510680735d3c79a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23939393939393938,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3d00b302a3967bb4fd844fa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"f3d00b302a3967bb4fd844fa","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23939393939393938},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","model_id":"m-0510680735d3c79a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"f2a1e415bd21c2ba8da14232a2f8ef65","model_id":"m-0510680735d3c79a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38585858585858585,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2f05ded322a1e6ed62ec7230"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"2f05ded322a1e6ed62ec7230","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38585858585858585},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"7d8a5af4b7693c52e2886778a85fed14","model_id":"m-0510680735d3c79a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4184839044652129,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b83b3553237e75b1f760428a"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"b83b3553237e75b1f760428a","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4184839044652129},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"c30f5d7bd4e32b6c25d7581895fd498c","model_id":"m-0510680735d3c79a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4131313131313131,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f683af1ba0a311b4de5be6d7"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"f683af1ba0a311b4de5be6d7","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4131313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","model_id":"m-0510680735d3c79a","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","model_id":"m-0510680735d3c79a","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0}],"official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profiles":[{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":1064,"numerator":64025},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":400},"family":"FrontierMath","row_ids":["2e3935359cea7e94134e77b2b6c16a83"],"score":57.14},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":75},"family":"GPQA Diamond","row_ids":["06022591029d9808b7b2c118e22277aa"],"score":75.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1350},"family":"Terminal-Bench-Science","row_ids":["5555c117d96b482e6d1bb0b304cbbefd"],"score":71.05},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":75},"family":"HLE","row_ids":["52db35ab0a7c574170e312e2a679c471"],"score":37.5}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v2","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","peer_count":39,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2}],"score":60.17,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":6,"exact":{"denominator":756,"numerator":37175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":500},"family":"APEX-Agents","row_ids":["100759b239352aa8328473780a99c2b5"],"score":71.43},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":400},"family":"GDPval-AA","row_ids":["8aeb7f071b6b514a7531446104ba33a1"],"score":44.44},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":200},"family":"Terminal-Bench","row_ids":["340dbce590b6aaaa8e1dc907ce4c05ef"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":125},"family":"AutomationBench","row_ids":["29e83d6eda99c0b01ee6e4a2561a6131"],"score":62.5},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":250},"family":"AA-Briefcase","row_ids":["02e903393ff93c77cfd706e931424b8b"],"score":27.78},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":200},"family":"MCP Atlas","row_ids":["938965e9ee08fb381a111173123203b2"],"score":22.22}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","included":true,"model_id":"m-0510680735d3c79a","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"2.1","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"4.0","peer_count":49,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v1.1","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8}],"score":49.17,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["ff5c1b3b8cf739706c1637caa41bfd3b"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1.1","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":1,"numerator":70},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"IFBench","row_ids":["3cf0fe91d3392b3852bb737856d256d6"],"score":40.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"GDP.pdf","row_ids":["5f7d8e688eab4e132d7071fdcec04c3f"],"score":100.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7}],"score":70.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":81,"numerator":5600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1100},"family":"Humanity Sixth Sense","row_ids":["8a450f924e67dc0b617d0ac58b1154ec"],"score":40.74},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":100},"family":"MMMU-Pro","row_ids":["267266c10eb3ebd28421d19f9b4fb8da"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"Chartography","row_ids":["c10b5e0559e0dd2ee1b376a9b0908819"],"score":66.67}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","peer_count":28,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9}],"score":69.14,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v2","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","peer_count":39,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","included":true,"model_id":"m-0510680735d3c79a","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"2.1","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"4.0","peer_count":49,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v1.1","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1.1","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","peer_count":28,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9}],"result_updated_at":"2026-09-29","score":60.75,"score_exact":{"denominator":344736,"numerator":20942839}}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"source_id":"openai/gpt-5.6-sol"},"m-0cd6b9e2d04f2c30":{"id":"m-0cd6b9e2d04f2c30","in_window":true,"name":"MiniCPM-V 4.6 1.3B","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":4,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":200},"family":"GPQA Diamond","row_ids":["c0db3282b4d173e68108771ffab82d19"],"score":3.39}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":3.39,"percentile_exact":{"denominator":59,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505}],"score":3.39,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":81,"numerator":100},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":100},"family":"GDPval-AA","row_ids":["32822274d56779c973cdcf3ccb12841f"],"score":1.23}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","peer_count":82,"percentile":1.23,"percentile_exact":{"denominator":81,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32}],"score":1.23,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"IFBench","row_ids":["e359de5ff096ac45c506a2baeb714ed7"],"score":0.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","peer_count":17,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551}],"score":0.0,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["7a567c9ea42d069b6cee3f34d322f5a9"],"score":0.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087}],"score":0.0,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":3.39,"percentile_exact":{"denominator":59,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","peer_count":82,"percentile":1.23,"percentile_exact":{"denominator":81,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","peer_count":17,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087}],"result_updated_at":null,"score":1.45,"score_exact":{"denominator":19116,"numerator":27625}}],"release_date":"2026-05-11","release_status":"published","release_verified":true,"source_id":"published/aa:minicpm-v-4-6-1-3b"},"m-0f7beea535abe0d6":{"id":"m-0f7beea535abe0d6","in_window":true,"name":"GPT-6.1 Sol","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d8d7274643c06b1a4ca112516b5792d9","model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":55.05,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"55.05%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":55.05},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f471d6e6533a0fc57044f213abd6c638","model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.850867052023121,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol-xhigh:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.850867052023121},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ba5d48e63b6204ebb5ad5eeadc76d418","model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.849132947976879,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol-high:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.849132947976879},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93746ce5b5a4350f9a12357e0a25f021","model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.839306358381503,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.839306358381503},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"55aa6d4f6d5c6c7039dc96c4b71c62e1","model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.83121387283237,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.83121387283237},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c51e4301e97e2734934dea53147ceb17","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1503.18,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1503.18},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6593896b3fbc00e60057e5fc83a0a4f0","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1468.52,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1468.52},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"503cd1a91f0f9d61632b123804df72fe","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1360.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1360.05},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37f540a6b34f9167b8de80e62424e45b","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1114.11,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1114.11},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43baa5ad2bffd18cda84f98b75d1d9ef","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6106666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6106666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0df4f5e9300e0d3a48fd6da07be51d2f","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.604,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.604},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f6fa274a901cc8dd20273af150c376e8","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.596},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95e456c7ee138e980560c08a90e0993c","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5863333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5863333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5cab5b2d874cb1fc27278bb8106d188d","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5196917808219178,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5196917808219178},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0608e71f47700ecac1426a1eea35d8e5","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5101010101010102,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5101010101010102},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"164d93e1ab11fa36ccef8c571d728a72","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.518976897689769,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.518976897689769},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8217dc87c4bd87cdc380a6aa4f1915b9","model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5012087026591459,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5012087026591459},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eac07b96c9ef79b7cc3356af0e88f7e","model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6657372141364344,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6657372141364344},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"570af2e51ac1cfc6154821482075a6a1","model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6449523732184466,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6449523732184466},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f8da2d3065e1826eb5ff8d8dca39499","model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6257350970435687,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6257350970435687},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ef7ad893c1714560eedcd7e2512dd35","model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5257088964747538,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5257088964747538},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ed264fc620044b8034b628539a63a072","model_id":"m-0f7beea535abe0d6","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.504761904761905,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-sol-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.504761904761905},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"707523b862efcb5d5f26677f23bfc017","model_id":"m-0f7beea535abe0d6","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.461904761904762,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-sol-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.461904761904762},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3f46bfda0775d3cdad5ddcf3394cf0f0","model_id":"m-0f7beea535abe0d6","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-sol-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c9ff6ecd964f033f20c34b6056abebda","model_id":"m-0f7beea535abe0d6","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.580952380952381,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-sol:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.580952380952381},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"33c15cf84256669aa69db60f413d6c22","model_id":"m-0f7beea535abe0d6","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.10952380952381,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-sol-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10952380952381},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e7ae1c14e0c9a4734b23aeae1079d58d","model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.385190909090909,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385190909090909},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"53b1bc2b78f3b152ed3916454df9a4f7","model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.36737297979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36737297979798},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9c082a143ac96e90b028f52d32b6ca1d","model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.333875757575758,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333875757575758},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"376a5e305d7986a188da9508896e601c","model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.201560858585859,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201560858585859},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba485c54bfc3a9d9cd36245ddafa34dd","model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.558,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5cc4b769351c2b5bd1699bc6df6d43","model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.519,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.519},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b4456a739cafe5c4ce1c21cbc7d31ce","model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.46,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.46},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7756eb954e080dc150407c7373d2f060","model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.382,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.382},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ccb37e73d86012a2e66e10678f5a8dbf","model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.318,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"848c930869b8d887ae35b3d57b4d09e0","model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.32,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c56dbd23d518d73e3bcdf682719ff251","model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.3,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"999b73f18a34bc1b86a7d623311c8428","model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.27,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6b5413928aee57cd272db71af132696","model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.54040404040404,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.54040404040404},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2799eee03b19a4ba7622e0bbcdc0787","model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.515151515151515,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.515151515151515},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9845a9479be0e3f5cafd550110575aef","model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.47979797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47979797979798},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7430e040011afcc017c249536ee1cd76","model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.308080808080808,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.308080808080808},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"458137fca9e267a9d61ef28bbd5c46c5","model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a439433bd255bc7be2bbe07c50bb162f","model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.685,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.685},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9edf287c404a36cfc04e0aec73b57f9","model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.67,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.67},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46c59f3b29bb0c6f9f08b0be77e5d67e","model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.625},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d8ae9a229631dda35d477d203d13fc2","model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.8,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dbbe9b1c99bd7d4f9855d2c8c2cdad33","model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1520.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1520.45},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a34433357a196f5e4efc5ee0d4b9312c","model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1459.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1459.05},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec42b8d14ada118567ac45e582c65e24","model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1321.98,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1321.98},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-29","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":3958475},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":100},"family":"FrontierMath","row_ids":["cacd9412ff6ba035ea9e826693b09343"],"score":100.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1850},"family":"Terminal-Bench-Science","row_ids":["77175d98c0074eb53a919fa599931946"],"score":97.37},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5900},"family":"HLE","row_ids":["9165499121e5e198448d4536365d9c16"],"score":93.65},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":5000},"family":"CritPt","row_ids":["96d62eb6c5d675afc73f8103f6dbc627"],"score":98.04}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","peer_count":39,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","peer_count":64,"percentile":93.65,"percentile_exact":{"denominator":63,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","peer_count":52,"percentile":98.04,"percentile_exact":{"denominator":51,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695}],"score":97.26,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4542615,"numerator":363697727},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2000},"family":"GDPval-AA","row_ids":["eaa5c4189ba4b51426be70de02910e04"],"score":74.07},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":272},"family":"Terminal-Bench","row_ids":["d04f5531bb77f209689af5059ec9471d"],"score":90.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1575},"family":"AutomationBench","row_ids":["4bd5c8368df638bd0c064dd06c306d48"],"score":82.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6400},"family":"AA-Briefcase","row_ids":["c7f025daa6b0e96ab52dc7df7e3bb96a"],"score":83.12},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1600},"family":"AA-AnalystAgent","row_ids":["9b9634fd2b2367e076a491de1e4f02e6"],"score":69.57}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","peer_count":82,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","peer_count":76,"percentile":90.67,"percentile_exact":{"denominator":3,"numerator":272},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":78,"percentile":83.12,"percentile_exact":{"denominator":77,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","peer_count":24,"percentile":69.57,"percentile_exact":{"denominator":23,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5}],"score":80.06,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":57425},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4500},"family":"FrontierCode","row_ids":["319a73d6f8fba5d605260942d6b8e601"],"score":91.84},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1150},"family":"Vals Code Migration","row_ids":["e330fa37e52713545b5fec3f612240b8"],"score":88.46}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","peer_count":50,"percentile":91.84,"percentile_exact":{"denominator":49,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","peer_count":27,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32}],"score":90.15,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":12153775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3675},"family":"GDP.pdf","row_ids":["f0c0c3073a678bbcb3f5f855da9adff8"],"score":96.71},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1600},"family":"SimpleQA Verified","row_ids":["509c52385e7c5559b4414a3e4bdb93cc"],"score":94.12},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"MultiChallenge","row_ids":["2a39ee205e56064a7ebeddb7db41c3ee"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3400},"family":"AA-Omniscience","row_ids":["ad20786a83f6b0b25b2f134cb2f6c095","8c01a1619e7c484574ab42025543f92a"],"score":64.15}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075}],"score":88.74,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":42175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2500},"family":"Humanity Sixth Sense","row_ids":["3a0983451c53206a0fa5061a863b8a25"],"score":92.59},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1550},"family":"MMMU-Pro","row_ids":["9a17a04ad8c7b9671f59a85153438a32"],"score":91.18}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","peer_count":28,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335}],"score":91.88,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","peer_count":39,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","peer_count":64,"percentile":93.65,"percentile_exact":{"denominator":63,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","peer_count":52,"percentile":98.04,"percentile_exact":{"denominator":51,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","peer_count":82,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","peer_count":76,"percentile":90.67,"percentile_exact":{"denominator":3,"numerator":272},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":78,"percentile":83.12,"percentile_exact":{"denominator":77,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","peer_count":24,"percentile":69.57,"percentile_exact":{"denominator":23,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","peer_count":50,"percentile":91.84,"percentile_exact":{"denominator":49,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","peer_count":27,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","peer_count":28,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335}],"result_updated_at":"2026-09-29","score":89.3,"score_exact":{"denominator":496604728620,"numerator":44346438312619}}],"release_date":"2026-09-29","release_status":"published","release_verified":true,"source_id":"openai/gpt-6.1-sol"},"m-12d2cbffeaae72d5":{"id":"m-12d2cbffeaae72d5","in_window":true,"name":"Granite 4.2 3B","observed_records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9},{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA-Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"130da9dac3f554ec6bf0afc38a355782","model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA-Diamond; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"62f4969756e5cb55e0903043","raw_value":55.9,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"55.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row22:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA-Diamond","source_record_id":"twelve-minicpm2-card:table0:row22:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":55.9},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":20,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":26078,"numerator":848075},"families":[{"benchmark_version":"2026","exact":{"denominator":3,"numerator":200},"family":"AIME","row_ids":["9bcefe5681f1de0d56a93cb5ca7a6411"],"score":66.67},{"benchmark_version":"500-item subset","exact":{"denominator":1,"numerator":100},"family":"MATH-500","row_ids":["48e7cd14f60fe5daf4244ca5ca1c19b2"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":600},"family":"GPQA Diamond","row_ids":["7e93a07b82dfd690e1c2b1309992332e"],"score":10.17},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d1c19809bb6f7e59cd9274e82fb1b4b5"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":0},"family":"HLE","row_ids":["20cab07f9ffc43544259f41d417e323c"],"score":0.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":50},"family":"CritPt","row_ids":["e23c22def1e7c8613f7559601fcfa581"],"score":0.98}],"family_count":6,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":10.17,"percentile_exact":{"denominator":59,"numerator":600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","peer_count":64,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":32.52,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":6162156,"numerator":60705805},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":125},"family":"tau3-Banking","row_ids":["9e32e2abcb9a8a594fb5e6aa3bcbd501"],"score":4.81},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":100},"family":"GDPval-AA","row_ids":["a9600d8248b3ea039e7316ffc060f044"],"score":3.7},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["c56ed05cd995d4a9a967c8a8edd7c23f"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":50},"family":"AutomationBench","row_ids":["de96f31cb9e7df111014b360185583e3"],"score":2.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":100},"family":"AA-Briefcase","row_ids":["caf840132854029b414465a52c19889d"],"score":1.3},{"benchmark_version":"4","exact":{"denominator":3,"numerator":100},"family":"BFCL","row_ids":["c9019a0cb657ee535803619a061368cf"],"score":33.33}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","peer_count":82,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":2.63,"percentile_exact":{"denominator":19,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":78,"percentile":1.3,"percentile_exact":{"denominator":77,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2}],"score":9.85,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":147,"numerator":5125},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["d294135a5f841df5749d74dee1cdb9e7"],"score":3.06},{"benchmark_version":"release_v6","exact":{"denominator":3,"numerator":200},"family":"LiveCodeBench","row_ids":["2f001b0be2d50fef9306d2b3f9f1569b"],"score":66.67}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9}],"score":34.86,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":477,"numerator":25225},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"IFBench","row_ids":["a7d850e0b3d6527713784858f859aa3f"],"score":66.67},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":200},"family":"Multi-IF","row_ids":["cf2e1f5dc39c9244f47a835c7f33569d"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"GDP.pdf","row_ids":["d9c7d28e2d9651e546a84191d6ab8ec7"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"IFEval","row_ids":["55369a687291e4012c741d522a892ee5"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":50},"family":"MMLU-Pro","row_ids":["8519642492aa86c8bdd1ad71a0f84568"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["432ed3bb705952b8c4dcd7c697936b37","7f11b1b4129f0979e12158aab3657dcc"],"score":33.96}],"family_count":6,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509}],"score":52.88,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":10.17,"percentile_exact":{"denominator":59,"numerator":600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","peer_count":64,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","peer_count":82,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":2.63,"percentile_exact":{"denominator":19,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":78,"percentile":1.3,"percentile_exact":{"denominator":77,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509}],"result_updated_at":null,"score":31.27,"score_exact":{"denominator":41274330401304,"numerator":1290617742549175}}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"source_id":"published/aa:granite-4-2-3b"},"m-161fc8601724c921":{"id":"m-161fc8601724c921","in_window":true,"name":"LFM2.5-8B-A1B","observed_records":[{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"4a80dc819df2f6636f58a1f4b11dd6c3","model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MATH500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"c40c5d7e6b8ae0ee0d913954","raw_value":88.76,"research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"88.76","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row7:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH500","source_record_id":"twelve-liquid8-card:table4:row7:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":88.76},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","model_id":"m-161fc8601724c921","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:28.652094+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-4451f79b3b503fb7","id":"497c9e05097ae6a248d4bb0700554017","model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b4387a964480dd771e528294","raw_value":91.84,"research_artifact_sha256":"8929a0523866355a99f3b750544896b93b1a812dd12eac6366cf11a33487ca45","response_sha256":"aa914e1990e6fe4bcce5c0b09e1338fbd76a801d47d17c1a31b29363220a1726","result_date":"2026-05-28","source_cell":"91.84","source_id":"s-2f340d11d74c0e6a7077","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-release:table0:row4:column2"},"source_name":"www.liquid.ai","source_protocol_key":"official:www.liquid.ai:IFEval:IFEval","source_record_id":"twelve-liquid8-release:table0:row4:column2","source_url":"https://www.liquid.ai/blog/lfm2-5-8b-a1b","unit":"%","value":91.84},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:28.652094+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-4451f79b3b503fb7","id":"7cc79eabbe19443b1f0528073b241cc7","model_id":"m-161fc8601724c921","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"32ad823060486b238d5bdec3","raw_value":56.47,"research_artifact_sha256":"8929a0523866355a99f3b750544896b93b1a812dd12eac6366cf11a33487ca45","response_sha256":"aa914e1990e6fe4bcce5c0b09e1338fbd76a801d47d17c1a31b29363220a1726","result_date":"2026-05-28","source_cell":"56.47","source_id":"s-2f340d11d74c0e6a7077","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-release:table0:row5:column2"},"source_name":"www.liquid.ai","source_protocol_key":"official:www.liquid.ai:IFBench:IFBench","source_record_id":"twelve-liquid8-release:table0:row5:column2","source_url":"https://www.liquid.ai/blog/lfm2-5-8b-a1b","unit":"%","value":56.47},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:28.652094+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-4451f79b3b503fb7","id":"e65ec69e51e864d98638aea829dce447","model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"82bd3283d06704264e6b8e6b","raw_value":79.93,"research_artifact_sha256":"8929a0523866355a99f3b750544896b93b1a812dd12eac6366cf11a33487ca45","response_sha256":"aa914e1990e6fe4bcce5c0b09e1338fbd76a801d47d17c1a31b29363220a1726","result_date":"2026-05-28","source_cell":"79.93","source_id":"s-2f340d11d74c0e6a7077","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-release:table0:row6:column2"},"source_name":"www.liquid.ai","source_protocol_key":"official:www.liquid.ai:Multi-IF:Multi-IF","source_record_id":"twelve-liquid8-release:table0:row6:column2","source_url":"https://www.liquid.ai/blog/lfm2-5-8b-a1b","unit":"%","value":79.93},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:27:28.652094+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-4451f79b3b503fb7","id":"cfcd3e6161d8c598825fd1053293931a","model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MATH500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"25715d0caf8b8efba4b9c159","raw_value":88.76,"research_artifact_sha256":"8929a0523866355a99f3b750544896b93b1a812dd12eac6366cf11a33487ca45","response_sha256":"aa914e1990e6fe4bcce5c0b09e1338fbd76a801d47d17c1a31b29363220a1726","result_date":"2026-05-28","source_cell":"88.76","source_id":"s-2f340d11d74c0e6a7077","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-release:table0:row7:column2"},"source_name":"www.liquid.ai","source_protocol_key":"official:www.liquid.ai:MATH-500:MATH500","source_record_id":"twelve-liquid8-release:table0:row7:column2","source_url":"https://www.liquid.ai/blog/lfm2-5-8b-a1b","unit":"%","value":88.76},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:28.652094+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-4451f79b3b503fb7","id":"caf7edeca636c003a7b36b6d88908822","model_id":"m-161fc8601724c921","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b555325944fa61bc3081767f","raw_value":48.5,"research_artifact_sha256":"8929a0523866355a99f3b750544896b93b1a812dd12eac6366cf11a33487ca45","response_sha256":"aa914e1990e6fe4bcce5c0b09e1338fbd76a801d47d17c1a31b29363220a1726","result_date":"2026-05-28","source_cell":"48.50","source_id":"s-2f340d11d74c0e6a7077","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-release:table0:row10:column2"},"source_name":"www.liquid.ai","source_protocol_key":"official:www.liquid.ai:BFCL:BFCLv4","source_record_id":"twelve-liquid8-release:table0:row10:column2","source_url":"https://www.liquid.ai/blog/lfm2-5-8b-a1b","unit":"%","value":48.5},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8},{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","model_id":"m-161fc8601724c921","original_version_label":"2026","profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9a125a53ca950d873e4577684bf386d1","model_id":"m-161fc8601724c921","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":51.0,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"51.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":51.0},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA-Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0cba5bb1737ffc1f4e69a7db0bd621bc","model_id":"m-161fc8601724c921","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA-Diamond; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"62f4969756e5cb55e0903043","raw_value":51.3,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"51.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row22:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA-Diamond","source_record_id":"twelve-minicpm2-card:table0:row22:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":51.3},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"61ff3d8dc14ad46a6ba4cce95c802615","model_id":"m-161fc8601724c921","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":49.2,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"49.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":49.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":531,"numerator":13000},"families":[{"benchmark_version":"2026","exact":{"denominator":3,"numerator":100},"family":"AIME","row_ids":["072b96c67ae19a20ee48a971117995c0"],"score":33.33},{"benchmark_version":"500-item subset","exact":{"denominator":3,"numerator":100},"family":"MATH-500","row_ids":["158a7a0bcbb4fdee990dff1e880fb156"],"score":33.33},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":400},"family":"GPQA Diamond","row_ids":["249b24c9d6325f053552fb911f7f4a84"],"score":6.78}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","included":true,"model_id":"m-161fc8601724c921","original_version_label":"2026","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":6.78,"percentile_exact":{"denominator":59,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313}],"score":24.48,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"4","exact":{"denominator":1,"numerator":0},"family":"BFCL","row_ids":["602ba59a6dfe14df55137c17751e7666"],"score":0.0}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","included":true,"model_id":"m-161fc8601724c921","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5}],"score":0.0,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"release_v6","exact":{"denominator":1,"numerator":0},"family":"LiveCodeBench","row_ids":["445a31daff84293600dd4a2900cbf97d"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":48,"numerator":775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":125},"family":"IFBench","row_ids":["950f9e2f9181b2c99ee5f8151838596d"],"score":31.25},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"Multi-IF","row_ids":["5b52e92a8499178ddfa84c801132f2f4"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":100},"family":"IFEval","row_ids":["0c588d45b55a074bc2915de39e2cf5ff"],"score":33.33},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"MMLU-Pro","row_ids":["0129fdb380d08e0198547b85719a77cd"],"score":0.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1}],"score":16.15,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","included":true,"model_id":"m-161fc8601724c921","original_version_label":"2026","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":6.78,"percentile_exact":{"denominator":59,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","included":true,"model_id":"m-161fc8601724c921","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1}],"result_updated_at":null,"score":10.39,"score_exact":{"denominator":38232,"numerator":397175}}],"release_date":"2026-05-28","release_status":"published","release_verified":true,"source_id":"published/aa:lfm2-5-8b-a1b"},"m-16a8dce2a6e7167a":{"id":"m-16a8dce2a6e7167a","in_window":true,"name":"Gemini 3.7 Flash","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14}],"official_evidence":{"artifact_sha256":"b68cd4829f86d5c5a5cafdbccf2e60b6956a5bfc04074d538c1dd7d5c4a9a559","basis":"official_release_page","date":"2026-08-13","sha256":"b68cd4829f86d5c5a5cafdbccf2e60b6956a5bfc04074d538c1dd7d5c4a9a559","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash/"},"primary_profile_id":"native","profiles":[{"benchmark_count":6,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":50},"family":"Terminal-Bench-Science","row_ids":["174d2abbe0e35fdfac2a047ae2fa2208"],"score":50.0}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","peer_count":39,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14}],"score":50.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":72,"numerator":2375},"families":[{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":325},"family":"Terminal-Bench","row_ids":["d7c99e26bef72e8691defabf85dff136"],"score":27.08},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":350},"family":"AA-Briefcase","row_ids":["d43d801b58cb460394b97c15079fc130"],"score":38.89}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","peer_count":49,"percentile":27.08,"percentile_exact":{"denominator":12,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365}],"score":32.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"run 0.1.200","exact":{"denominator":2,"numerator":125},"family":"MirrorCode","row_ids":["02fcbee985f3da4b53d62ce022f2a290"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1400},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1400},"family":"SimpleQA Verified","row_ids":["ab75c5704ac20fa5dc1c93828558d2ce"],"score":82.35}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692}],"score":82.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":27,"numerator":2000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2000},"family":"Humanity Sixth Sense","row_ids":["e0d9bd2a6ca55a94058a8fdfafb19d29"],"score":74.07}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","peer_count":28,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8}],"score":74.07,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","peer_count":39,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","peer_count":49,"percentile":27.08,"percentile_exact":{"denominator":12,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","peer_count":28,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8}],"result_updated_at":"2026-10-21","score":57.12,"score_exact":{"denominator":14688,"numerator":839045}}],"release_date":"2026-08-13","release_status":"published","release_verified":true,"source_id":"google/gemini-3.7-flash"},"m-1d98b8595cf1927d":{"id":"m-1d98b8595cf1927d","in_window":true,"name":"Granite 4.2 8B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":28072925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":750},"family":"GPQA Diamond","row_ids":["80f85d325d6536dbd1603ecf4678090b"],"score":12.71},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["10a24da4314ddad1e5487011d9ce22be"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":200},"family":"HLE","row_ids":["1c035cc3ccd2789e7031fcc995e521f7"],"score":3.17},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":50},"family":"CritPt","row_ids":["8e1dd162a62fd85fef38ef1d97e4e9bc"],"score":0.98}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","peer_count":64,"percentile":3.17,"percentile_exact":{"denominator":63,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":8.54,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":18519664},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":175},"family":"tau3-Banking","row_ids":["a1e714634bbaa4b289e4726c4bb2c9e9"],"score":13.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1000},"family":"GDPval-AA","row_ids":["5be65e1004d9e6f482549aefe65a50a4"],"score":12.35},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["f435dfe97ff078f0f04715c2a33b7561"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":275},"family":"AutomationBench","row_ids":["e3a197e716166eb9edc4c46f49e52295"],"score":14.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":500},"family":"AA-Briefcase","row_ids":["2e097f3022508687e5c06477d01a5253"],"score":6.49}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":13.46,"percentile_exact":{"denominator":13,"numerator":175},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","peer_count":82,"percentile":12.35,"percentile_exact":{"denominator":81,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":78,"percentile":6.49,"percentile_exact":{"denominator":77,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41}],"score":12.02,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["6a0e7ddb2b9af667874b506e526b6d54"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":35525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":25},"family":"GDP.pdf","row_ids":["8eaabce7efa57428ec82adf89ba50dbc"],"score":1.32},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["4a7b9f5169709891b5917ee76e2a8e1a","5f020104c70b0364755cf0f561bff098"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647}],"score":17.64,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","peer_count":64,"percentile":3.17,"percentile_exact":{"denominator":63,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":13.46,"percentile_exact":{"denominator":13,"numerator":175},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","peer_count":82,"percentile":12.35,"percentile_exact":{"denominator":81,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":78,"percentile":6.49,"percentile_exact":{"denominator":77,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647}],"result_updated_at":null,"score":10.31,"score_exact":{"denominator":41274330401304,"numerator":425645973328915}}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"source_id":"ibm/granite-4.2-8b"},"m-1dcd13c2fce7d97b":{"id":"m-1dcd13c2fce7d97b","in_window":true,"name":"GPT-6.1 Terra","observed_records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6.1 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6b7dbc78828ec370d1445c253fb081f8","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":49.49,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"49.49%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row11","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":49.49},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"883caa20fb65b10b1b2df1a59575f69a","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.760115606936416,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760115606936416},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"0a9b9ded6fc0729100a56e6e71e37fdb","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bff9aec545c4bccb6f4ba623a000b02f","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.82485549132948,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.82485549132948},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d606a0b7672e2bbccd0464b1f272cf8d","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra-xhigh:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8238c3538921ccb179c31b64149ac668","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra-high:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1ec48f79a034e68fac381a130ff38268","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1225.85,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra-non-reasoning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1225.85},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a0e8941d112d5a0488fba3ac5368da8","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1190.07,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1190.07},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1b032180d6e45191f35f33260bf0190","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1030.56,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1030.56},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49a6347dbc84e646ff6e6ad7f1092bb1","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1333.47,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1333.47},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85907fa4d2c372369829d06684b448cb","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1263.74,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1263.74},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d28028745784d8c7000837c5fecd06b4","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4073333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-non-reasoning:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4073333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"bb18dd6aef9ef1dc30f32803de89de30","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47433333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47433333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0748e9152af905b251d0fe1a391ec1d4","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45566666666666666,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45566666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d8d47bcde74d7b9f2ed507ab01b7099c","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48033333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a8009fa29192e53947f64108cfd7fca","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4726666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4726666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"047a2ba4c16dee77598156833b2b21a4","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.22890888638920137,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-non-reasoning:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22890888638920137},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e761a75e7201ca9671dbf9e13e0897f","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6036778693722258,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036778693722258},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11f82142c1ef9184340a501cec984d78","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5842008573178199,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5842008573178199},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b90afae005dae1ca509397e6a5f2806b","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.632456703014753,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632456703014753},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"584d1fa4ea5a10f383cd3e5a3c0d1252","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6302149178255373,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6302149178255373},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5b02730bb566527cdac4413c8c9ea6af","model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5104424383087566,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra-non-reasoning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5104424383087566},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ce4d2de161c96d98f01b13d8046ab7cf","model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5780171778957812,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5780171778957812},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda811901bcee22723151f9536f96373","model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5624828754079192,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5624828754079192},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a0ae681b4c6de22b8e081176c2e82e7","model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5992855956353923,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5992855956353923},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4ec986b7fed5970273c9200687534fdf","model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.584605963496433,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.584605963496433},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1cfff6605bff3020759752a91d868cf0","model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.242857142857143,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra-non-reasoning:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.242857142857143},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"05d8ba27d5acba049f3086f1b96761fb","model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.247619047619048,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247619047619048},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1d1c001f360a13753eb6497752bdb574","model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0761904761904762,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0761904761904762},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cbb775e1df07a24e836a17d00b5214ac","model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.419047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.419047619047619},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5bb69e0ca82d4c45497b76c8e17ca7f","model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.333333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6de46b3787c0913d303fc365bce4b8d9","model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.333510606060606,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra-non-reasoning:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333510606060606},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aba4b59ea9724a0875727e22d8f920e9","model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.345359343434343,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.345359343434343},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f42b9148f859b931a03c590eda466176","model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.244852272727273,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.244852272727273},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1822e649b86eb233861ceec54919caf0","model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.368595202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.368595202020202},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b61c189833d9ab9ba8abd0b2c80356ae","model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.37457904040404,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37457904040404},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0f7b48cc02e94c43770d54ada21a4338","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.145,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-terra-non-reasoning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e85b4ef3cccfd0ec6841edff67807576","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.473,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-terra:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.473},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0d7a58cf052d1c4319c269366b78c92b","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.313,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-terra-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbcb1d28d1b8756bac4c6390684e96e3","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.249,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-terra-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.249},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ae492457aeed0a8044231b33e3816e3e","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.423,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-terra-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"096c218278789f1f9ee2afc1f8fef2b2","model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.354,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-terra-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.354},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b48be3cf7afc847acb7a93d1daa106f","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.232,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra-non-reasoning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e7b46084efd7cf1417a7e45fbbfa01ac","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b6c41c4eab4297eef3ee5b2baa0dafa","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.256,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.256},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b106a379be115e886f08407740c20833","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.266,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.266},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"51a826cd09ef8e67bb45c58abd9989d7","model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.28,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e94c0f5e59481de097ed13e23555ebf","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.267676767676768,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra-non-reasoning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.267676767676768},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9bda15c03a0261cd993cc0968e3ae75","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.323232323232323,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.323232323232323},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2fe675b61c73809778478a1763c5465d","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.151515151515152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.151515151515152},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc4d3bdb33a8fbf05f078078098440ce","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.464646464646465,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464646464646465},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"61d34e3aaa9da4faa4ceaa3d50c25a30","model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.419191919191919,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.419191919191919},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"dee1af74d2a02441d5a020e070932bb1","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra-non-reasoning:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"531cbbae99ac83e13be3bad033f3ca22","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0998734878cc273150388b19cd3bafa","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"88d491ab3289410979becd137db02a65","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.605,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.605},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"559b1a0f915626c7d43054c183a845dc","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.615,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.615},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"88ad6b79782ece2f005f5bd1c28b4edb","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1286.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-non-reasoning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra-non-reasoning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1286.1},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b66bc31908f62ffabe3017f3af39ed3","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1324.7,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1324.7},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2aba9c7eae6f18407bbee7d4d6fd3de8","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1203.75,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1203.75},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"963f90fbcb149ad855be710f3141b385","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1414.87,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1414.87},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"13086fb5e07b39e55477afdbe8f73560","model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1352.74,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1352.74}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":31824,"numerator":2630125},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":800},"family":"FrontierMath","row_ids":["428bcb6f2312f54ea131660ec5510f0c"],"score":88.89},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1200},"family":"Terminal-Bench-Science","row_ids":["3a64ac33a8e0bec7c9bbb47064e6200b"],"score":92.31},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":225},"family":"HLE","row_ids":["9c15f08ff3f1d95aac1ab4460a6a090b"],"score":56.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4750},"family":"CritPt","row_ids":["624fa427f469bd9c8115ca8491b3f8a8"],"score":93.14}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","peer_count":53,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68}],"score":82.65,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":17835809},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5600},"family":"GDPval-AA","row_ids":["b0fdd931f8b81700b478f4129dec593b"],"score":69.14},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":268},"family":"Terminal-Bench","row_ids":["35604c8a1315056deeba470a2491f933"],"score":89.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1450},"family":"AutomationBench","row_ids":["1b6722ff89e791988385d9c44034ffe3"],"score":76.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5100},"family":"AA-Briefcase","row_ids":["07c0ef81ba075e768f25e1177f4ce283"],"score":66.23}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","peer_count":82,"percentile":69.14,"percentile_exact":{"denominator":81,"numerator":5600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","peer_count":76,"percentile":89.33,"percentile_exact":{"denominator":3,"numerator":268},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":78,"percentile":66.23,"percentile_exact":{"denominator":77,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36}],"score":75.25,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":4700},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4700},"family":"FrontierCode","row_ids":["bf241286cb096d94e2aa92889567c092"],"score":95.92}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","peer_count":50,"percentile":95.92,"percentile_exact":{"denominator":49,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505}],"score":95.92,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":9124175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1775},"family":"GDP.pdf","row_ids":["5a11c32835ae26c4dc93dd63135b1a2a"],"score":93.42},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1150},"family":"SimpleQA Verified","row_ids":["52a58ef3bf54eb42e218e3c094aec598"],"score":67.65},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"MultiChallenge","row_ids":["c65e268d91030372566c3ec5c7a1ab64"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3600},"family":"AA-Omniscience","row_ids":["43cafc16779c4f9a017fe8ae9ccdcb81","5ac877348d0fb5ad33d4ed61b4154929"],"score":67.92}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129}],"score":66.62,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1450},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1450},"family":"MMMU-Pro","row_ids":["cd4afdb9196fc47218c2c1085364d08a"],"score":85.29}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555}],"score":85.29,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","peer_count":53,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","peer_count":82,"percentile":69.14,"percentile_exact":{"denominator":81,"numerator":5600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","peer_count":76,"percentile":89.33,"percentile_exact":{"denominator":3,"numerator":268},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":78,"percentile":66.23,"percentile_exact":{"denominator":77,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","peer_count":50,"percentile":95.92,"percentile_exact":{"denominator":49,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555}],"result_updated_at":"2026-10-26","score":80.51,"score_exact":{"denominator":621835486272,"numerator":50065762899577}}],"release_date":"2026-10-26","release_status":"published","release_verified":true,"source_id":"openai/gpt-6.1-terra"},"m-21b38c9ea818aebe":{"id":"m-21b38c9ea818aebe","in_window":true,"name":"Muse Wonder Full","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"dbbbd82ee29a731632cb4d28c2efdeea","model_id":"m-21b38c9ea818aebe","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.14,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"37.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":37.14},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc581128088f8e6a2ca283411c75c69c","model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder-full:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":5967,"numerator":520600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["0caa25aeefa10afc49082f57afb1971b"],"score":84.62},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":800},"family":"HLE","row_ids":["4e26762c5adc39cb8336491e6db427a8"],"score":88.89},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1500},"family":"CritPt","row_ids":["f2c7d3b6f687d90b18b5ecae64664442"],"score":88.24}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","peer_count":64,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","peer_count":52,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565}],"score":87.25,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3792096,"numerator":357306625},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7900},"family":"GDPval-AA","row_ids":["fa97f8f99150ab638ce3d789e781973c"],"score":97.53},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":2125},"family":"Terminal-Bench","row_ids":["07f1cdf7bbcb0da17ca1fb2ddf289fbe"],"score":88.54},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1775},"family":"AutomationBench","row_ids":["9d53e1ef4d2e9c319b7f0dacff98d83c"],"score":93.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7500},"family":"AA-Briefcase","row_ids":["4f4f6c3a22e75b41412d7850bc4b9cf5"],"score":97.4}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","peer_count":82,"percentile":97.53,"percentile_exact":{"denominator":81,"numerator":7900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","peer_count":49,"percentile":88.54,"percentile_exact":{"denominator":24,"numerator":2125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":78,"percentile":97.4,"percentile_exact":{"denominator":77,"numerator":7500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32}],"score":94.22,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":637,"numerator":51225},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4300},"family":"FrontierCode","row_ids":["36981a7ad16ac52e29b5d161b5c3a617"],"score":87.76},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":950},"family":"Vals Code Migration","row_ids":["3d507b92bee70e08544b9eb65b20267a"],"score":73.08}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","peer_count":50,"percentile":87.76,"percentile_exact":{"denominator":49,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","peer_count":27,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79}],"score":80.42,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":53,"numerator":4825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"GDP.pdf","row_ids":["b8e5730fd93ac21867d62c37b5d60e40"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4350},"family":"AA-Omniscience","row_ids":["a92d6d21339d4ee97a6fa90820847da1","211e11b4d985ead9c997be1fc30d7961"],"score":82.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418}],"score":91.04,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","peer_count":64,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","peer_count":52,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","peer_count":82,"percentile":97.53,"percentile_exact":{"denominator":81,"numerator":7900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","peer_count":49,"percentile":88.54,"percentile_exact":{"denominator":24,"numerator":2125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":78,"percentile":97.4,"percentile_exact":{"denominator":77,"numerator":7500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","peer_count":50,"percentile":87.76,"percentile_exact":{"denominator":49,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","peer_count":27,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418}],"result_updated_at":null,"score":88.51,"score_exact":{"denominator":5596519376448,"numerator":495344146904675}}],"release_date":"2026-10-12","release_status":"published","release_verified":true,"source_id":"meta/muse-wonder-full"},"m-25ccaf97cf82a5e0":{"id":"m-25ccaf97cf82a5e0","in_window":true,"name":"Ring-2.6-1T","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-08","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["7c25d00a19511ccc369639cf5605291e"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":183027373},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":500},"family":"tau3-Banking","row_ids":["3cd21f0f8001e029d7b19208e29ba099"],"score":38.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1700},"family":"GDPval-AA","row_ids":["2a88f9d5ec82d3f04004cf3da53914a5"],"score":20.99},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["7e0782bec5efa3fc907fbba656122a79"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":125},"family":"AutomationBench","row_ids":["bfaca7bcdefcec6c5d1daac93dc0d456"],"score":6.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1600},"family":"AA-Briefcase","row_ids":["d8f9d30bc51b5ac382358dbda4a533e5"],"score":20.78}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","peer_count":82,"percentile":20.99,"percentile_exact":{"denominator":81,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":78,"percentile":20.78,"percentile_exact":{"denominator":77,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95}],"score":23.76,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":12084,"numerator":238625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":75},"family":"IFBench","row_ids":["3080cdd2a4ddf6bfb1a2fed6805cd96e"],"score":18.75},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":375},"family":"GDP.pdf","row_ids":["5ecbc58b2efe8e4c7ad00b00fd409205"],"score":19.74},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1100},"family":"AA-Omniscience","row_ids":["f5ca62e67f09a75d5d460c81d0851fa4","2c99a44c194b274fedb14bdd3896ace9"],"score":20.75}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393}],"score":19.75,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","peer_count":82,"percentile":20.99,"percentile_exact":{"denominator":81,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":78,"percentile":20.78,"percentile_exact":{"denominator":77,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393}],"result_updated_at":null,"score":34.1,"score_exact":{"denominator":33720858171,"numerator":1150018852748}}],"release_date":"2026-05-08","release_status":"published","release_verified":true,"source_id":"published/aa:ring-2-6-1t"},"m-26bf3900b3ceb955":{"id":"m-26bf3900b3ceb955","in_window":true,"name":"K2 Horizon 3.7B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":117351,"numerator":1674775},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1100},"family":"GPQA Diamond","row_ids":["c96afdd76392eb5d2c9b157b68b7de9e"],"score":18.64},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d7eec79b902bf9a304fdfc120e9aa124"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":350},"family":"CritPt","row_ids":["9fef9d533791c4fe2ca276032af2b566"],"score":6.86}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":18.64,"percentile_exact":{"denominator":59,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01}],"score":14.27,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":42819149},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":550},"family":"tau3-Banking","row_ids":["7b39b47195e57aecded4286410a368f3"],"score":42.31},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2300},"family":"GDPval-AA","row_ids":["eada0c68a6bf5686fb53166d5b815ba4"],"score":28.4},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["585ae90ea798ef43916073acc85d178e"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":575},"family":"AutomationBench","row_ids":["eb41669f3fb02b0d8fee80ec8c840979"],"score":30.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1900},"family":"AA-Briefcase","row_ids":["ed41f889fc4e7786b137bd6b76a3da51"],"score":24.68}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","peer_count":82,"percentile":28.4,"percentile_exact":{"denominator":81,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":30.26,"percentile_exact":{"denominator":19,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","peer_count":78,"percentile":24.68,"percentile_exact":{"denominator":77,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97}],"score":27.79,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":600},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":600},"family":"FrontierCode","row_ids":["908a5049e650c9f2841b87a654313ae7"],"score":12.24}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","peer_count":50,"percentile":12.24,"percentile_exact":{"denominator":49,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798}],"score":12.24,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":625},"family":"GDP.pdf","row_ids":["e8733c3c5f23f9a2dee5ae8306fe34a6"],"score":16.45}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"score":16.45,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":18.64,"percentile_exact":{"denominator":59,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","peer_count":82,"percentile":28.4,"percentile_exact":{"denominator":81,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":30.26,"percentile_exact":{"denominator":19,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","peer_count":78,"percentile":24.68,"percentile_exact":{"denominator":77,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","peer_count":50,"percentile":12.24,"percentile_exact":{"denominator":49,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"result_updated_at":null,"score":18.06,"score_exact":{"denominator":97345118871,"numerator":1758167195585}}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"source_id":"published/aa:k2-horizon-3-7b"},"m-2c518178af26d1d4":{"id":"m-2c518178af26d1d4","in_window":true,"name":"North Mini Code","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","model_id":"m-2c518178af26d1d4","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","model_id":"m-2c518178af26d1d4","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","model_id":"m-2c518178af26d1d4","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-09","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1095276,"numerator":17204225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2000},"family":"GPQA Diamond","row_ids":["eebaf08b0e483d418dfe3796053cb2a9"],"score":33.9},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["b878256e74c2c14b269834f44b853cea"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":100},"family":"HLE","row_ids":["19f9062134531b120e5a3beb1b827183"],"score":4.76},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":350},"family":"CritPt","row_ids":["1ce53974b2fd41332a137498132cd83b"],"score":6.86}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":33.9,"percentile_exact":{"denominator":59,"numerator":2000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","peer_count":64,"percentile":4.76,"percentile_exact":{"denominator":21,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01}],"score":15.71,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":93810898},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":125},"family":"tau3-Banking","row_ids":["a7dc3e52deea9234fc190c75c4007564"],"score":9.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":500},"family":"GDPval-AA","row_ids":["72fabae9540ce0210203fd32c61249bb"],"score":6.17},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["c32f27bfb7b7fcf6e0fb4b8f49b0314a"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":175},"family":"AutomationBench","row_ids":["427f2b62a677b7b14b79a216cd5cce2a"],"score":9.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":300},"family":"AA-Briefcase","row_ids":["f3f72292d6ff516b21ceaec7da17a866"],"score":3.9}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":9.62,"percentile_exact":{"denominator":13,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","peer_count":82,"percentile":6.17,"percentile_exact":{"denominator":81,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":9.21,"percentile_exact":{"denominator":19,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":78,"percentile":3.9,"percentile_exact":{"denominator":77,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45}],"score":12.18,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":400},"family":"FrontierCode","row_ids":["e2d037eb56a1707481cab5a658eb199f"],"score":8.16}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"2.0","peer_count":50,"percentile":8.16,"percentile_exact":{"denominator":49,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101}],"score":8.16,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3021,"numerator":62900},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"IFBench","row_ids":["65c9161268988be090ca4411568e9e7b"],"score":37.5},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":375},"family":"GDP.pdf","row_ids":["96fc6325ca8f16276f80447a5bb8a745"],"score":9.87},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":800},"family":"AA-Omniscience","row_ids":["2d5bfdd87d275a4c6177037eb5a3ff85","30b64e451b40a0bb848f166f0f44b6dd"],"score":15.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037}],"score":20.82,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":33.9,"percentile_exact":{"denominator":59,"numerator":2000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","peer_count":64,"percentile":4.76,"percentile_exact":{"denominator":21,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":9.62,"percentile_exact":{"denominator":13,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","peer_count":82,"percentile":6.17,"percentile_exact":{"denominator":81,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":9.21,"percentile_exact":{"denominator":19,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":78,"percentile":3.9,"percentile_exact":{"denominator":77,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"2.0","peer_count":50,"percentile":8.16,"percentile_exact":{"denominator":49,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037}],"result_updated_at":null,"score":14.19,"score_exact":{"denominator":3174948492408,"numerator":45043683141617}}],"release_date":"2026-06-09","release_status":"published","release_verified":true,"source_id":"cohere/north-mini-code-1-0"},"m-2cd37409a483435a":{"id":"m-2cd37409a483435a","in_window":true,"name":"Claude Sonnet 5","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","model_id":"m-2cd37409a483435a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8ab92ff9c703946a72390a40d74174e5","model_id":"m-2cd37409a483435a","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":10.3,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"10.3%","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row2:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-4e3ca8a95d36:table0:row2:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":10.3},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","model_id":"m-2cd37409a483435a","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"1454e68858e26e7f25431416e9f35b82","model_id":"m-2cd37409a483435a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1359.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1359","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1359.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","model_id":"m-2cd37409a483435a","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"890ff0f1982cbf940e7560095d25c3d1","model_id":"m-2cd37409a483435a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.24646464646464647,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25ed1c0f160c101313b8b03f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"25ed1c0f160c101313b8b03f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.24646464646464647},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"bf11fdb01a5abb04e9237f452f243fe0","model_id":"m-2cd37409a483435a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.42323232323232324,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9c2cf6199ae2109c00de1e74"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9c2cf6199ae2109c00de1e74","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.42323232323232324},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","model_id":"m-2cd37409a483435a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52d4a201d4e85f74412406318a4344d6","model_id":"m-2cd37409a483435a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.35454545454545455,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e13f1e993e8d9faa1fe4cd20"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e13f1e993e8d9faa1fe4cd20","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.35454545454545455},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"e4cafd5fbbdde21911996a705a220ab2","model_id":"m-2cd37409a483435a","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2909090909090909,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f08fcdeb661758c2d6aeda86"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"f08fcdeb661758c2d6aeda86","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2909090909090909},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"822c2fa500974ad4cbdc6d8832137e95","model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.337,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:42.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6ef2dabb93c56c7defda8883"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"6ef2dabb93c56c7defda8883","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.337},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","model_id":"m-2cd37409a483435a","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"official_evidence":{"artifact_sha256":"78c9b451825b0f421c2e785a5934defd9662507c3913b2f7c2c0df07917ba424","basis":"official_release_page","date":"2026-06-30","sha256":"78c9b451825b0f421c2e785a5934defd9662507c3913b2f7c2c0df07917ba424","url":"https://www.anthropic.com/news/claude-sonnet-5"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":850},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["01d54ae905ebb77f621f6cbf30eb0930"],"score":44.74}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"score":44.74,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":96,"numerator":2525},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":100},"family":"GDPval-AA","row_ids":["fcb5b9889977ea6f8824ae1c412b655a"],"score":11.11},{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":175},"family":"Terminal-Bench","row_ids":["bbf181bad55ef3d8ecad294914ed6b67"],"score":21.88},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":500},"family":"AA-Briefcase","row_ids":["14b2d91b066241cbb4ba0914b53ae0a4"],"score":55.56},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":50},"family":"OSWorld","row_ids":["cdb864bc0c75bd1fb1f8c43c173d544c"],"score":16.67}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"v1.1","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0}],"score":26.3,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["494d4b4ce09501b6feb0dbb80233b707"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":150},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":150},"family":"SimpleQA Verified","row_ids":["0c6f13f860966ddef45cc04368fd5127"],"score":8.82}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329}],"score":8.82,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":250},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":500},"family":"Humanity Sixth Sense","row_ids":["cf4a733238c901c9de673ff40841fce0"],"score":18.52},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["a4e7fdc6e1f2295b627a11265f28666d"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","peer_count":28,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6}],"score":9.26,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"v1.1","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","peer_count":28,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6}],"result_updated_at":"2026-09-28","score":32.95,"score_exact":{"denominator":1116288,"numerator":36782095}}],"release_date":"2026-06-30","release_status":"published","release_verified":true,"source_id":"anthropic/claude-sonnet-5"},"m-2d46f5cc91cc119d":{"id":"m-2d46f5cc91cc119d","in_window":true,"name":"JT-4.1 Flash 236B A21B","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9c7dcf7c0575b062e703557bce8a40b","model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1128.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1128.36},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-09","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":3100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3100},"family":"GPQA Diamond","row_ids":["012394f027c8418fe233e5f26b558306"],"score":52.54}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":52.54,"percentile_exact":{"denominator":59,"numerator":3100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545}],"score":52.54,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":118503,"numerator":6339076},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4700},"family":"GDPval-AA","row_ids":["87c313476c9a420ad27b41db7d0250bf"],"score":58.02},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":154},"family":"Terminal-Bench","row_ids":["8ad3da1ee13871ee0748612b7b0f2202"],"score":51.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1050},"family":"AutomationBench","row_ids":["f3bd4620346190964b5a6a49765359b6"],"score":55.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3800},"family":"AA-Briefcase","row_ids":["6f61748806659fcf7e221477d42b6563"],"score":49.35}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","peer_count":82,"percentile":58.02,"percentile_exact":{"denominator":81,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","peer_count":78,"percentile":49.35,"percentile_exact":{"denominator":77,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44}],"score":53.49,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":725},"family":"GDP.pdf","row_ids":["c038eca51adef7673b4117a8c612883d"],"score":38.16}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086}],"score":38.16,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":100},"family":"MMMU-Pro","row_ids":["90c2787f5b2a0845d431bc6af09c8c25"],"score":5.88}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746}],"score":5.88,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":52.54,"percentile_exact":{"denominator":59,"numerator":3100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","peer_count":82,"percentile":58.02,"percentile_exact":{"denominator":81,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","peer_count":78,"percentile":49.35,"percentile_exact":{"denominator":77,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746}],"result_updated_at":null,"score":43.41,"score_exact":{"denominator":475434036,"numerator":20638975985}}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"source_id":"published/aa:jt-4-1-flash-236b-a21b"},"m-2f22b07da3af863c":{"id":"m-2f22b07da3af863c","in_window":true,"name":"Muse Spark 1.2","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","model_id":"m-2f22b07da3af863c","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06}],"official_evidence":{"artifact_sha256":"92a7e106c1f5836b322f76df67124f4f08c92c8dbe8cfa67e78698bd91fb844f","basis":"official_release_page","date":"2026-08-05","sha256":"92a7e106c1f5836b322f76df67124f4f08c92c8dbe8cfa67e78698bd91fb844f","url":"https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2"},"primary_profile_id":"native","profiles":[{"benchmark_count":4,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":45.45,"evaluator_count":2,"exact":{"denominator":144,"numerator":6425},"families":[{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":275},"family":"Terminal-Bench","row_ids":["bd6d00a622ee8666fe845f21e2f0650f"],"score":11.46},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":700},"family":"AA-Briefcase","row_ids":["2a4003430f96eaca6cf7fbdac62ade55"],"score":77.78}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154}],"score":44.62,"weight":25},{"effective_weight_percent":36.36,"evaluator_count":1,"exact":{"denominator":17,"numerator":1200},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1200},"family":"SimpleQA Verified","row_ids":["aa61e58d244812473e1af5a8764b4e07"],"score":70.59}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844}],"score":70.59,"weight":20},{"effective_weight_percent":18.18,"evaluator_count":1,"exact":{"denominator":27,"numerator":400},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":400},"family":"Humanity Sixth Sense","row_ids":["b1617164f055d3d72ddaf46f34da89d8"],"score":14.81}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","peer_count":28,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5}],"score":14.81,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","peer_count":28,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5}],"result_updated_at":"2026-08-27","score":48.64,"score_exact":{"denominator":80784,"numerator":3929575}}],"release_date":"2026-08-05","release_status":"published","release_verified":true,"source_id":"meta/muse-spark-1.2"},"m-3535a3e7891af5ef":{"id":"m-3535a3e7891af5ef","in_window":true,"name":"DeepSeek V4 Flash 0731","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"cb8519de408cafe4f74e189f2d0ff6f2","model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":7.0,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"7.0","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":7.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"5aa8713728b0b23d443020c1571a6c86","model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":54.4,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.4","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"official_evidence":{"artifact_sha256":"3f80af72789dd0dd52b0bd5de471e2df2bc33c6eb443bf98eaeecc4dec6afbf4","basis":"official_release_page","date":"2026-07-31","sha256":"3f80af72789dd0dd52b0bd5de471e2df2bc33c6eb443bf98eaeecc4dec6afbf4","url":"https://api-docs.deepseek.com/updates/"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":57,"numerator":1375},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":50},"family":"GPQA Diamond","row_ids":["021a82f4f7be85e4b693852ce483ef03"],"score":16.67},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["5a0b6bca7e87cdf2b14ffe16a27cf187"],"score":31.58}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":24.12,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":54,"numerator":1975},"families":[{"benchmark_version":"4.0","exact":{"denominator":2,"numerator":75},"family":"Terminal-Bench","row_ids":["6bb0c331de392cf1dc35a1de5de30317"],"score":37.5},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["25c531e52c90edae0ad06f8953351733"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":650},"family":"AA-Briefcase","row_ids":["239fbb8019bce72c55d5b79e4b9186e5"],"score":72.22}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","peer_count":49,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516}],"score":36.57,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3,"numerator":200},"families":[{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":200},"family":"DeepSWE","row_ids":["72c84440fb70eb9d01c3241fc4d3e551"],"score":66.67}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4}],"score":66.67,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"IFBench","row_ids":["bb7586265eb9f4ee440c1de9eb72708e"],"score":33.33},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":200},"family":"SimpleQA Verified","row_ids":["f310a7a1d6e944a484a86a2598b55b31"],"score":11.76}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636}],"score":22.55,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","peer_count":49,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636}],"result_updated_at":"2026-08-27","score":36.69,"score_exact":{"denominator":313956,"numerator":11517775}}],"release_date":"2026-07-31","release_status":"published","release_verified":true,"source_id":"deepseek/deepseek-v4-flash-0731"},"m-35c4f287acc76c95":{"id":"m-35c4f287acc76c95","in_window":true,"name":"Inkling Small","observed_records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","model_id":"m-35c4f287acc76c95","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","model_id":"m-35c4f287acc76c95","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","model_id":"m-35c4f287acc76c95","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","model_id":"m-35c4f287acc76c95","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fbda3169b181acd989a4524a5ca42c7f","model_id":"m-35c4f287acc76c95","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling-small:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-30","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":365092,"numerator":15288225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4300},"family":"GPQA Diamond","row_ids":["70d4757947a89f4f143b905fc208d5b4"],"score":72.88},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["569c655345f8f6c2c6fa591358c80baa"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":800},"family":"HLE","row_ids":["3c34699bd1ded8eb4f916044a6a266ef"],"score":38.1},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2000},"family":"CritPt","row_ids":["0bfa1e021f0f3d031ecf9fe9a047c9e1"],"score":39.22}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":72.88,"percentile_exact":{"denominator":59,"numerator":4300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","peer_count":64,"percentile":38.1,"percentile_exact":{"denominator":21,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","peer_count":52,"percentile":39.22,"percentile_exact":{"denominator":51,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185}],"score":41.87,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":755891136,"numerator":23344506125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":525},"family":"tau3-Banking","row_ids":["b9bbbaa648034918f0dabaffc06da3bd"],"score":40.38},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1300},"family":"GDPval-AA","row_ids":["fd7e6b8b2243df420c1327f60f09d326"],"score":48.15},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":25},"family":"Terminal-Bench","row_ids":["2f300304d9d00cf74194cf028a1e2a05"],"score":1.04},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":425},"family":"AutomationBench","row_ids":["441205ca0d2d0b6e786d210564b45527"],"score":22.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2500},"family":"AA-Briefcase","row_ids":["b87f8c47a158b30537b60e5453c5b4d9"],"score":32.47},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":550},"family":"EnterpriseOps-Gym","row_ids":["fe3cd8a1335112d2ed6891e37698119e"],"score":61.11},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":100},"family":"MCP Atlas","row_ids":["72495a708e9f00af5b2d17370ac07628"],"score":11.11},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":700},"family":"AA-AnalystAgent","row_ids":["6890e628f5988bd5a798a680de5a69d4"],"score":30.43}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":40.38,"percentile_exact":{"denominator":13,"numerator":525},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","peer_count":82,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"4.0","peer_count":49,"percentile":1.04,"percentile_exact":{"denominator":24,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":22.37,"percentile_exact":{"denominator":19,"numerator":425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":78,"percentile":32.47,"percentile_exact":{"denominator":77,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","peer_count":24,"percentile":30.43,"percentile_exact":{"denominator":23,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275}],"score":30.88,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":9675},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1300},"family":"FrontierCode","row_ids":["066fdac9a6f7fc061fce980c5a691276"],"score":26.53},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":50},"family":"Vals Code Migration","row_ids":["17347953383a652f8a8fabeabf1d07e6"],"score":3.85}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"2.0","peer_count":50,"percentile":26.53,"percentile_exact":{"denominator":49,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1.1","peer_count":27,"percentile":3.85,"percentile_exact":{"denominator":13,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43}],"score":15.19,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":6042,"numerator":170825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["17242a9341b3cc9173ea459907ce9455"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":0},"family":"SimpleQA Verified","row_ids":["62e95547f0f7073efcdd1420b882e4a5"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1950},"family":"AA-Omniscience","row_ids":["c31ed6f5932783f10dbc8e7e910353cc","01245c922f864bcfcfa4ad1336dea016"],"score":36.79}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433}],"score":28.27,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":400},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":400},"family":"MMMU-Pro","row_ids":["93ce1046100f7e346ed0a342303a4604"],"score":23.53}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665}],"score":23.53,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":72.88,"percentile_exact":{"denominator":59,"numerator":4300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","peer_count":64,"percentile":38.1,"percentile_exact":{"denominator":21,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","peer_count":52,"percentile":39.22,"percentile_exact":{"denominator":51,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":40.38,"percentile_exact":{"denominator":13,"numerator":525},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","peer_count":82,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"4.0","peer_count":49,"percentile":1.04,"percentile_exact":{"denominator":24,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":22.37,"percentile_exact":{"denominator":19,"numerator":425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":78,"percentile":32.47,"percentile_exact":{"denominator":77,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","peer_count":24,"percentile":30.43,"percentile_exact":{"denominator":23,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"2.0","peer_count":50,"percentile":26.53,"percentile_exact":{"denominator":49,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1.1","peer_count":27,"percentile":3.85,"percentile_exact":{"denominator":13,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665}],"result_updated_at":"2026-08-27","score":29.23,"score_exact":{"denominator":1125107673161472,"numerator":32892305162137165}}],"release_date":"2026-07-30","release_status":"published","release_verified":true,"source_id":"thinkingmachines/inkling-small"},"m-36451e3f44503d37":{"id":"m-36451e3f44503d37","in_window":true,"name":"Nemotron 3.5 Super VL","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","model_id":"m-36451e3f44503d37","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","model_id":"m-36451e3f44503d37","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","model_id":"m-36451e3f44503d37","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","model_id":"m-36451e3f44503d37","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","model_id":"m-36451e3f44503d37","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","model_id":"m-36451e3f44503d37","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","model_id":"m-36451e3f44503d37","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","model_id":"m-36451e3f44503d37","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":41769,"numerator":1004075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["42505f6b2ea9972da7c497ad966aebac"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1600},"family":"HLE","row_ids":["30beb0844cc09771243cbe9dfff243cd"],"score":25.4},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":500},"family":"CritPt","row_ids":["e6e2851045fbef435d3da74124d06523"],"score":29.41}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"Diamond","peer_count":64,"percentile":25.4,"percentile_exact":{"denominator":63,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","peer_count":52,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135}],"score":24.04,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":10773,"numerator":369385},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4000},"family":"GDPval-AA","row_ids":["3c9b5c57e54a04df82602451589cb092"],"score":49.38},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["01e4bcd3555fdf1970b9b729ff28db44"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":600},"family":"AutomationBench","row_ids":["ae26dd2850d89ddaa6b644b249715c9f"],"score":31.58},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":300},"family":"AA-Briefcase","row_ids":["b0a10dcacf57ed9fc64b8506d23f5ae1"],"score":42.86}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2.1","peer_count":82,"percentile":49.38,"percentile_exact":{"denominator":81,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":78,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76}],"score":34.29,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1600},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1600},"family":"FrontierCode","row_ids":["35228bba34f7d198ea6ae7b85139d6dc"],"score":32.65}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"2.0","peer_count":50,"percentile":32.65,"percentile_exact":{"denominator":49,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141}],"score":32.65,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":79475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":675},"family":"GDP.pdf","row_ids":["26277d899631407b16323e0f427ca69b"],"score":35.53},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["41481c594a0d6acdcc3cd6076cd49938","1c32657b7924b6111d38086436775d4a"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161}],"score":39.46,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"Diamond","peer_count":64,"percentile":25.4,"percentile_exact":{"denominator":63,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","peer_count":52,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2.1","peer_count":82,"percentile":49.38,"percentile_exact":{"denominator":81,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":78,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"2.0","peer_count":50,"percentile":32.65,"percentile_exact":{"denominator":49,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161}],"result_updated_at":null,"score":32.23,"score_exact":{"denominator":7949601387,"numerator":256194148075}}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"source_id":"published/aa:nemotron-3-5-super-vl"},"m-383f40728e3e5259":{"id":"m-383f40728e3e5259","in_window":true,"name":"Hy4 preview","observed_records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"531a4a7604ebef6b0434ef211d42dfb1","model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"90afbbf55be36ca70eb520d2","raw_value":37.1,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","model_id":"m-383f40728e3e5259","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"7cc7a31c8754c107bbefba22dfda40ad","model_id":"m-383f40728e3e5259","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.1,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","model_id":"m-383f40728e3e5259","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","model_id":"m-383f40728e3e5259","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","model_id":"m-383f40728e3e5259","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Hy4 preview / Terminus 2","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"cf97b6bb89d81346120bcb8bd77a6f88","model_id":"m-383f40728e3e5259","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Terminus 2. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"d56f8d18d1a15b35d44ddeda","raw_value":1.43,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Terminus 2. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","model_id":"m-383f40728e3e5259","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36}],"official_evidence":{"artifact_sha256":"676214cf8f2ac19b63d1293254a5e1920f5c5b758546dd77d7bbb0d132589ea3","basis":"official_release_page","date":"2026-08-28","sha256":"676214cf8f2ac19b63d1293254a5e1920f5c5b758546dd77d7bbb0d132589ea3","url":"https://www.tencent.com/zh-hk/tencent-releases-and-open-sources-tencent-hy4-preview/"},"primary_profile_id":"native","profiles":[{"benchmark_count":6,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7,"numerator":400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":400},"family":"GPQA Diamond","row_ids":["11f1be50c8d04272ce6fde6b77e7d32d"],"score":57.14}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"reference_revision":"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3}],"score":57.14,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":3,"exact":{"denominator":21,"numerator":550},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"APEX-Agents","row_ids":["7155f49a20e936c90be455db70c3df42"],"score":33.33},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":50},"family":"Terminal-Bench","row_ids":["54a707812fad1f7aeac89c3fb45c1b29"],"score":16.67},{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":200},"family":"AutomationBench","row_ids":["f68fab6bf158dd2600b8426c509f20a3"],"score":28.57}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"4.0","peer_count":49,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"reference_revision":"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1}],"score":26.19,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":2,"exact":{"denominator":91,"numerator":3350},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":400},"family":"Vals Code Migration","row_ids":["8b32286773fba058c9bb87cdd6ac0c71"],"score":30.77},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":300},"family":"DeepSWE","row_ids":["49c0ea386ce0f4eee60fb157edec7d0d"],"score":42.86}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"1.1","peer_count":27,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3}],"score":36.81,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"reference_revision":"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"4.0","peer_count":49,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"reference_revision":"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"1.1","peer_count":27,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3}],"result_updated_at":null,"score":40.28,"score_exact":{"denominator":1911,"numerator":76975}}],"release_date":"2026-08-28","release_status":"published","release_verified":true,"source_id":"tencent/hy4-preview"},"m-3ad11f2b6253b364":{"id":"m-3ad11f2b6253b364","in_window":true,"name":"Nemotron 3.5 Bolt 74B-A9B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":474012,"numerator":12975565},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2600},"family":"GDPval-AA","row_ids":["a6f95ea45eef0a446f5d6ed10bf3ca1b"],"score":32.1},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["614dafc454b6d64e61b2f25573ec4717"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":625},"family":"AutomationBench","row_ids":["d9c233b856dc8ecc623629216c3b1c31"],"score":32.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2400},"family":"AA-Briefcase","row_ids":["e9c29563b5cc25446296aa38a0ecd4ae"],"score":31.17}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","peer_count":82,"percentile":32.1,"percentile_exact":{"denominator":81,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","peer_count":78,"percentile":31.17,"percentile_exact":{"denominator":77,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94}],"score":27.37,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":2,"numerator":25},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"GDP.pdf","row_ids":["b6eecb655a01ff6addbb4005e50d5c51"],"score":12.5}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028}],"score":12.5,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","peer_count":82,"percentile":32.1,"percentile_exact":{"denominator":81,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","peer_count":78,"percentile":31.17,"percentile_exact":{"denominator":77,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028}],"result_updated_at":null,"score":20.76,"score_exact":{"denominator":4266108,"numerator":88578425}}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"source_id":"published/aa:nemotron-3-5-bolt-74b-a9b"},"m-3b5ea6fdfa94bc08":{"id":"m-3b5ea6fdfa94bc08","in_window":true,"name":"GPT-6 Sol","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":152,"numerator":9325},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1500},"family":"Terminal-Bench-Science","row_ids":["f1163968f642895b63614f184b419fdc"],"score":78.95},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":175},"family":"HLE","row_ids":["3ff9ef5159d8b127a6433bdd9ec141a5"],"score":43.75}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","peer_count":39,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8}],"score":61.35,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":1235},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":200},"family":"GDPval-AA","row_ids":["8b095392eb68858c1d89a63205401933"],"score":22.22},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":75},"family":"Terminal-Bench","row_ids":["a142db33093dc7d6f056b8d8509f429e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":40},"family":"AA-Briefcase","row_ids":["ad451031d1d4a9ddfab5c379bfc5d852"],"score":40.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","peer_count":49,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"reference_revision":"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0}],"score":45.74,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["0c33bab4ff416863111d8ddbb4eb7f0b"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8}],"score":100.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1250},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1250},"family":"SimpleQA Verified","row_ids":["8a0eac6d7e9eaac51a5eeee920a35f4d"],"score":73.53}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607}],"score":73.53,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":970},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1400},"family":"Humanity Sixth Sense","row_ids":["e25930e774ad7abbc0f5c7ffcc567aa5"],"score":51.85},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"Chartography","row_ids":["961e2599d1347e8ad57a29581acdb50c"],"score":20.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","peer_count":28,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6}],"score":35.93,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","peer_count":39,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","peer_count":49,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"reference_revision":"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","peer_count":28,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6}],"result_updated_at":"2026-10-26","score":65.07,"score_exact":{"denominator":93024,"numerator":6053149}}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"source_id":"openai/gpt-6-sol"},"m-3c17c0d3ca7f1abe":{"id":"m-3c17c0d3ca7f1abe","in_window":true,"name":"Grok 4.6","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"89573f9726831597d277e7bf81d0aa47","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5383838383838384,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"66cc58962988fb8f8115f8d7"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"66cc58962988fb8f8115f8d7","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5383838383838384},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"c8f3354d85d54309fb4c0591bbf6f004","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5252525252525253,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7843bf6e2811a25bab4c95f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7843bf6e2811a25bab4c95f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5252525252525253},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"f5b415d0191854dbb0bdd66144bc7513","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5282828282828282,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88428f7983913b1e9f3c61a"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d88428f7983913b1e9f3c61a","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5282828282828282},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"af82111f1a2cd52920c9adf6ac00c757","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.489,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:19.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93eef47574ea50d45679ea83"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"93eef47574ea50d45679ea83","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.489},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43}],"official_evidence":{"artifact_sha256":"22a9c3dcee05d0c5a93b52c78e1c1e867278c4f39ebda057faec281332c01411","basis":"official_release_page","date":"2026-08-12","sha256":"22a9c3dcee05d0c5a93b52c78e1c1e867278c4f39ebda057faec281332c01411","url":"https://x.ai/news/grok-4-6"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":19,"numerator":1200},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1200},"family":"Terminal-Bench-Science","row_ids":["545c704b3f9b5b5c3a00fbe6211a666c"],"score":63.16}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","peer_count":39,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43}],"score":63.16,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":144,"numerator":6875},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":275},"family":"Terminal-Bench","row_ids":["620e1ee85f651d96e2f6e21d676ecc23"],"score":34.38},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":550},"family":"AA-Briefcase","row_ids":["45f7b80c7cc89c267cb8e0d74593b6e4"],"score":61.11}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202}],"score":47.74,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"run 0.1.200","exact":{"denominator":2,"numerator":125},"family":"MirrorCode","row_ids":["0621e61b8874eac3232c0422d1f71677"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111}],"score":62.5,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":17,"numerator":800},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":800},"family":"SimpleQA Verified","row_ids":["954b3bb3d889d2d8fa0f84b9a3223eaf"],"score":47.06}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493}],"score":47.06,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","peer_count":39,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493}],"result_updated_at":"2026-10-21","score":55.15,"score_exact":{"denominator":837216,"numerator":46174325}}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"source_id":"xai/grok-4.6"},"m-3fe24456aa661a33":{"id":"m-3fe24456aa661a33","in_window":true,"name":"GPT-5.5 Instant (June 2026)","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","model_id":"m-3fe24456aa661a33","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":164300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2500},"family":"GPQA Diamond","row_ids":["4ac219e06b407a4da2953a84f21a360f"],"score":42.37},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2900},"family":"HLE","row_ids":["5706c727b7c4831ffe377f7ebcc17796"],"score":46.03}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":42.37,"percentile_exact":{"denominator":59,"numerator":2500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","peer_count":64,"percentile":46.03,"percentile_exact":{"denominator":63,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146}],"score":44.2,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":293031139},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":250},"family":"tau3-Banking","row_ids":["c4bd6d87d676b473fe7bbadf8b7238e4"],"score":19.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1100},"family":"GDPval-AA","row_ids":["0bf079be27f6b7285cbdd993ff5b5a93"],"score":13.58},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":178},"family":"Terminal-Bench","row_ids":["4473f9853fd2b2892365909edf9631fb"],"score":59.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":975},"family":"AutomationBench","row_ids":["20e66aebafb2226fab822f3a7524d54c"],"score":51.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3600},"family":"AA-Briefcase","row_ids":["110052f3248350cfda8c6580116d3dd9"],"score":46.75}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","peer_count":82,"percentile":13.58,"percentile_exact":{"denominator":81,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"4.0","peer_count":76,"percentile":59.33,"percentile_exact":{"denominator":3,"numerator":178},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":51.32,"percentile_exact":{"denominator":19,"numerator":975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","peer_count":78,"percentile":46.75,"percentile_exact":{"denominator":77,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66}],"score":38.04,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":1450},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1450},"family":"GDP.pdf","row_ids":["9ab89470db83472d1b5d2ed28ebc6783"],"score":76.32}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202}],"score":76.32,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":42.37,"percentile_exact":{"denominator":59,"numerator":2500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","peer_count":64,"percentile":46.03,"percentile_exact":{"denominator":63,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","peer_count":82,"percentile":13.58,"percentile_exact":{"denominator":81,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"4.0","peer_count":76,"percentile":59.33,"percentile_exact":{"denominator":3,"numerator":178},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":51.32,"percentile_exact":{"denominator":19,"numerator":975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","peer_count":78,"percentile":46.75,"percentile_exact":{"denominator":77,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202}],"result_updated_at":null,"score":51.18,"score_exact":{"denominator":1272485214,"numerator":65122894901}}],"release_date":"2026-06-25","release_status":"published","release_verified":true,"source_id":"published/aa:gpt-5-5-instant-06-26"},"m-407d9206347b790d":{"id":"m-407d9206347b790d","in_window":true,"name":"MiniCPM5-2B","observed_records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1},{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","model_id":"m-407d9206347b790d","original_version_label":"2026","profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA-Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"13ab52d0b216c84bb8c4111e06baab4e","model_id":"m-407d9206347b790d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA-Diamond; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"62f4969756e5cb55e0903043","raw_value":70.2,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row22:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA-Diamond","source_record_id":"twelve-minicpm2-card:table0:row22:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.2},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","model_id":"m-407d9206347b790d","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","model_id":"m-407d9206347b790d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","model_id":"m-407d9206347b790d","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","model_id":"m-407d9206347b790d","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","model_id":"m-407d9206347b790d","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-07","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":16,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3717,"numerator":175250},"families":[{"benchmark_version":"2026","exact":{"denominator":1,"numerator":100},"family":"AIME","row_ids":["0f70d96aa863b344234e9f6fb914e9f2"],"score":100.0},{"benchmark_version":"500-item subset","exact":{"denominator":3,"numerator":200},"family":"MATH-500","row_ids":["1f0d5271a4aa3a380367a096ae7ae210"],"score":66.67},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1200},"family":"GPQA Diamond","row_ids":["d95e6a6f385be987c8949ca6952efa66"],"score":20.34},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":100},"family":"HLE","row_ids":["75182b4b31b9383c53f3f18d0e09c65d"],"score":1.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"2026","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":20.34,"percentile_exact":{"denominator":59,"numerator":1200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond","peer_count":64,"percentile":1.59,"percentile_exact":{"denominator":63,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038}],"score":47.15,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":18486468,"numerator":627714865},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":1225},"family":"tau3-Banking","row_ids":["dc8c6983f496e505270c3f1d2e08a1c1"],"score":47.12},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1600},"family":"GDPval-AA","row_ids":["f2ca1c75e7e337b37337e6b5eb3f5cd5"],"score":19.75},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["49929d3422a54a6833b2d573268b4f8f"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":225},"family":"AutomationBench","row_ids":["4a74a50d45014989b724a431c32f6c56"],"score":11.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":900},"family":"AA-Briefcase","row_ids":["8a20412103b559527bf3c1a6c26511e5"],"score":11.69},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["9fed535a16c1cf8e07c033c97e3a6783"],"score":100.0}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","included":true,"model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v2.1","peer_count":82,"percentile":19.75,"percentile_exact":{"denominator":81,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","included":true,"model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v1.1","peer_count":78,"percentile":11.69,"percentile_exact":{"denominator":77,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v4","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6}],"score":33.96,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"release_v6","exact":{"denominator":1,"numerator":100},"family":"LiveCodeBench","row_ids":["97029c45627dc4bbce9101042fef65e2"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1}],"score":100.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":171,"numerator":5495},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"IFBench","row_ids":["30e5129b448d30519f98acdcb4661feb"],"score":44.44},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":100},"family":"Multi-IF","row_ids":["ca424296b6739f797da253012aa25cc9"],"score":33.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":150},"family":"GDP.pdf","row_ids":["9ffb2abc9760a5a8e507be89d245bb8f"],"score":7.89},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"IFEval","row_ids":["c5c1fcc77312c8bfdb8f9a2733828c50"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":75},"family":"MMLU-Pro","row_ids":["fdfed7f2ef133f7ec35f427f9ee8f813"],"score":75.0}],"family_count":5,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8}],"score":32.13,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"2026","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":20.34,"percentile_exact":{"denominator":59,"numerator":1200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond","peer_count":64,"percentile":1.59,"percentile_exact":{"denominator":63,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","included":true,"model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v2.1","peer_count":82,"percentile":19.75,"percentile_exact":{"denominator":81,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","included":true,"model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v1.1","peer_count":78,"percentile":11.69,"percentile_exact":{"denominator":77,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v4","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8}],"result_updated_at":null,"score":51.89,"score_exact":{"denominator":19632629016,"numerator":1018776441535}}],"release_date":"2026-09-07","release_status":"published","release_verified":true,"source_id":"openbmb/minicpm5-2b"},"m-422e5ba87859dd55":{"id":"m-422e5ba87859dd55","in_window":true,"name":"Gemini 3.8 Flash","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"agents","direction":"higher_better","id":"60b632021bc45635455840f54b94c4f4","model_id":"m-422e5ba87859dd55","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"ac0f80ad7501f371bea4ff3e","raw_value":19.1,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row1:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench:Terminal-Bench 4.0","source_record_id":"gpt6-astra-official:table2:row1:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":19.1},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","model_id":"m-422e5ba87859dd55","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"reasoning","direction":"higher_better","id":"4523e0388c826e5440921d454fffec5e","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Gemini 3.8 Flash. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":95.3,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"95.3%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":95.3},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"ed92f5e3f147fe079ea8a940470282f6","model_id":"m-422e5ba87859dd55","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","model_id":"m-422e5ba87859dd55","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9f47dcf421f9d07c71bd917f2795e68c","model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.842196531791908,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.842196531791908},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e72423cc0ea9bb49329e0146a13850dd","model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.845086705202312,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845086705202312},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49a687c760081fcc9d21857eecec73d3","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1183.13,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1183.13},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9795339f2cd9e24671187cc6b1d0440","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1061.67,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1061.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77ace87006826d798f70dc4966eaa579","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49633333333333335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49633333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b51897039499e9fcbe335e5d65d153c","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4826666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f271222696cae9b647b5d0635cfd17a2","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48510919920582396,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48510919920582396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"47537605b70be0a787292544d2bafd50","model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.31894329896907214,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31894329896907214},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"850d47513974a9890ccc4c0cdd3a85ce","model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6088997392114149,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6088997392114149},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7249181f5f8d8d32dbb02a437152e80","model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3650940461486253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3650940461486253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ab146f114fc6f6b005fe452cdd897ee1","model_id":"m-422e5ba87859dd55","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.104761904761905,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-8-flash-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.104761904761905},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d203cb7570f8997ff9e7dbde81b27140","model_id":"m-422e5ba87859dd55","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-8-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9527a65a587ec475f195f02f8cd3b8c6","model_id":"m-422e5ba87859dd55","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.080952380952381,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-8-flash-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.080952380952381},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b8e90131f6bcd6c49a21c3a2932b1dc","model_id":"m-422e5ba87859dd55","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.164714393939394,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164714393939394},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","model_id":"m-422e5ba87859dd55","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"489131bc5de49fa007fc9387e16cde39","model_id":"m-422e5ba87859dd55","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0806391414141414,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0806391414141414},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"25441d91f50e6e749035c7559262480b","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.291,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-8-flash-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.291},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ec8a0a14f14573543e2e6dd2102670c","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.331,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-8-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.331},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5db1b060888870d1e32cac5146839ddd","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-8-flash-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b146a07e787e7a60ffe503d1736631e","model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b423aaf36506dd46f43cd4284bd6f9e1","model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e4a8a5066150dc5d684ec04e715afe95","model_id":"m-422e5ba87859dd55","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","model_id":"m-422e5ba87859dd55","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"100039733f13259411315ef088d09590","model_id":"m-422e5ba87859dd55","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.101010101010101},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47fbd95b8d167033cb09665c3782b98e","model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db768bc76d113f83c2369af327221b81","model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"99c2b783a1942b3a21b76c99dfb11641","model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1431.32,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1431.32},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f01d26441ea13fdbb60ea48a2bd9cbac","model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1316.07,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1316.07},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d89cd448a3fc0c0d307c8402bbf62e91","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash-medium:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca9c1febd80728f1c6244226144b28df","model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.92020202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash-low:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.92020202020202},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf80b93d53b3c4448b69a5e432495bad","model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.457731958762887,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-medium:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash-medium:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.457731958762887},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e231b9795fe1f2f485ffc92be4546b70","model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.331958762886598,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.331958762886598}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":114342,"numerator":7985875},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5800},"family":"GPQA Diamond","row_ids":["93a25d2afe298e1250966bfe70f4f70f"],"score":98.31},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1000},"family":"Terminal-Bench-Science","row_ids":["4c2b94e36c85b9d8f2e3d83490a34034"],"score":52.63},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":50},"family":"HLE","row_ids":["08dc502d8d4aeb42de93130ed556d391"],"score":50.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4000},"family":"CritPt","row_ids":["4e088fc1433335f9f212c9f9ffa8b974"],"score":78.43}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":98.31,"percentile_exact":{"denominator":59,"numerator":5800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","peer_count":39,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","peer_count":17,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","peer_count":52,"percentile":78.43,"percentile_exact":{"denominator":51,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465}],"score":69.84,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":5594589,"numerator":380137666},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1125},"family":"tau3-Banking","row_ids":["ceb2e04d7badea9a2869bdbce0586635"],"score":86.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5300},"family":"GDPval-AA","row_ids":["5e4d6ed857d58561b430218e5d2e3d99"],"score":65.43},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":64},"family":"Terminal-Bench","row_ids":["477576b238b66114c394a26ca4a20345"],"score":64.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"AutomationBench","row_ids":["44e1f8cce1da50e102dced9197ed9f4e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4300},"family":"AA-Briefcase","row_ids":["c6c37db45460e80aac980c7255d8ffed"],"score":55.84},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1400},"family":"AA-AnalystAgent","row_ids":["9cd6a63a2e9021c4431ddd906f3978b5"],"score":60.87}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":86.54,"percentile_exact":{"denominator":13,"numerator":1125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","peer_count":82,"percentile":65.43,"percentile_exact":{"denominator":81,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"4.0","peer_count":76,"percentile":64.0,"percentile_exact":{"denominator":1,"numerator":64},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":78,"percentile":55.84,"percentile_exact":{"denominator":77,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","peer_count":24,"percentile":60.87,"percentile_exact":{"denominator":23,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625}],"score":67.95,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":3822,"numerator":208475},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2600},"family":"FrontierCode","row_ids":["926e007cfcb9b2d7a75e59641d96bd9e"],"score":53.06},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":300},"family":"Vals Code Migration","row_ids":["59c32ec585e3e137a1a5f19c3bda0e98"],"score":23.08},{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":175},"family":"DeepSWE","row_ids":["e589efcb9de8d6dea5758589d780110f"],"score":87.5}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"2.0","peer_count":50,"percentile":53.06,"percentile_exact":{"denominator":49,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":27,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":9,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8}],"score":54.55,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":3835475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1525},"family":"GDP.pdf","row_ids":["1bf434a1cea1439ab199e26911f14113"],"score":80.26},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1450},"family":"SimpleQA Verified","row_ids":["f06072893cbefd23dc63a86432677691"],"score":85.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3100},"family":"AA-Omniscience","row_ids":["1cb384dc9fec91c5db3c1841d018dfc3","4b5e9db335123a5e12d2351e281e6182"],"score":58.49}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916}],"score":74.68,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":38950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2200},"family":"Humanity Sixth Sense","row_ids":["11851315d0cbcd3709f1ea2206b9b74b"],"score":81.48},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1500},"family":"MMMU-Pro","row_ids":["8bcd59ede7a9782bc9b8b3c169c19887"],"score":88.24}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","peer_count":28,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185}],"score":84.86,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":98.31,"percentile_exact":{"denominator":59,"numerator":5800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","peer_count":39,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","peer_count":17,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","peer_count":52,"percentile":78.43,"percentile_exact":{"denominator":51,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":86.54,"percentile_exact":{"denominator":13,"numerator":1125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","peer_count":82,"percentile":65.43,"percentile_exact":{"denominator":81,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"4.0","peer_count":76,"percentile":64.0,"percentile_exact":{"denominator":1,"numerator":64},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":78,"percentile":55.84,"percentile_exact":{"denominator":77,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","peer_count":24,"percentile":60.87,"percentile_exact":{"denominator":23,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"2.0","peer_count":50,"percentile":53.06,"percentile_exact":{"denominator":49,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":27,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":9,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","peer_count":28,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185}],"result_updated_at":"2026-09-29","score":68.78,"score_exact":{"denominator":18613913710392,"numerator":1280244857994287}}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"source_id":"google/gemini-3.8-flash"},"m-5553b7ca5f3e748a":{"id":"m-5553b7ca5f3e748a","in_window":true,"name":"GPT-6 Astra","observed_records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"agents","direction":"higher_better","id":"583c9d1e2da082747c42402376e2e5f7","model_id":"m-5553b7ca5f3e748a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":41.4,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":41.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"agents","direction":"higher_better","id":"3817ce25a9a6b0599640f57a9cca3571","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"ac0f80ad7501f371bea4ff3e","raw_value":57.9,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row1:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench:Terminal-Bench 4.0","source_record_id":"gpt6-astra-official:table2:row1:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":57.9},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"c754fd67237feab08d16e123eec898ee","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench Science 0.1; preserve page footnotes and task-specific settings","protocol_id":"6e315c8a4731951529a013b5","raw_value":64.6,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row1:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench-Science:Terminal-Bench Science 0.1","source_record_id":"gpt6-astra-official:table3:row1:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":64.6},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"e40148b85734dc0f9f8ce08ccb902b8d","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑6 Astra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":96.0,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"96.0%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":96.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench 4.0 (AA)","benchmark_version":"4.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"109d26a2b8ea6ca9e4a3d9999b7c5ce6","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench 4.0 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"6dfc6fa538421f2fee6d35c4","raw_value":59.1,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"59.1%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench 4.0 (AA):GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench:Terminal-Bench 4.0 (AA)","source_record_id":"carbon-image:Terminal-Bench 4.0 (AA):GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":59.1},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench Zapier","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"d28a3120853d4efe5840e7bb51950fc3","model_id":"m-5553b7ca5f3e748a","original_version_label":"current published developer protocol","profile_key":"native","protocol_description":"Official comparison image; AutomationBench Zapier; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"f658b6663efcee3dc42343bd","raw_value":41.4,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"41.4%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:AutomationBench Zapier:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:AutomationBench:AutomationBench Zapier","source_record_id":"carbon-image:AutomationBench Zapier:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":41.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science 0.1 (AA)","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"reasoning","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"11d354c28bc62a914ec4cd9b63089852","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench-Science 0.1 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"4f191cd5eb9caf57e48bc0e7","raw_value":63.3,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"63.3%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench-Science 0.1 (AA):GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench-Science:Terminal-Bench-Science 0.1 (AA)","source_record_id":"carbon-image:Terminal-Bench-Science 0.1 (AA):GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":63.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-6 Astra","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"83392963b6824da10354750a678d58cf","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Agentic coding Terminal-Bench 4.0¹; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":57.9,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"57.9%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row2:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-0dd3bed5c7dd:table0:row2:column4","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":57.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-6 Astra","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"c006e5f0249adc5c225cc895f365ee07","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1542.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1542","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column4","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1542.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-6 Astra","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"03836c4b9ecec508c6c4b2750f3aead4","model_id":"m-5553b7ca5f3e748a","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Business workflows AutomationBench²; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":41.4,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"41.4%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row10:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"official-0dd3bed5c7dd:table0:row10:column4","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":41.4},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-6 Astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"bf3d7c1e3fc2beed65427c4ff2fa2858","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Agentic scientific research Terminal-Bench-Science 0.1³; preserve page footnotes and task-specific settings","protocol_id":"1975ca8c09f87d3fc7597c12","raw_value":64.6,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"64.6%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row14:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1","source_record_id":"official-0dd3bed5c7dd:table0:row14:column4","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":64.6},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d139017f7e4a52c147ebd99f332ea3fc","model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86242774566474,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra-xhigh:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86242774566474},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b1f168b9c9a9a3fd93cdd96bcc1df5b7","model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846242774566474,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846242774566474},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"68d1d27d4144794872686bb043a6f355","model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.864161849710983,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra-high:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.864161849710983},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6e65945f05ff915a82469d93af8b0758","model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.850867052023121,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.850867052023121},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af24713a785067d91aa17b91f685e461","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1543.53,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1543.53},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"645b0e66a83b3b916bbaf8e56ef8f2e5","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1256.49,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1256.49},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"673148745f44386e570115e79caa5508","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1504.54,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1504.54},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"13fc060c43cf52741f7110daa7552c12","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1460.93,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1460.93},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad32258e07bf1831d0d650b1b69e9914","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6103333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6103333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"aa99fb17d630fb77127c41f9d5583005","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.586,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.586},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a44ffd00f5473404a16cc94af01c88c","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.611,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.611},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1b3b6f51887728d73a53f35084f410a1","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5963333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5963333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"381a774de2ce5ede748f16ede3110ce5","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5543199315654406,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5543199315654406},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9b6cb62304bd6d97b7c1050beddc070f","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5579710144927537,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5579710144927537},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e8392e8e492ff9990e86d0bc3cc385","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5749785775492716,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5749785775492716},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d66e00f9bd740ea79c41502fe3978a66","model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5483071841453344,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5483071841453344},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aeb332a491bc033ab35dcffb2d34d64b","model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6717778020215392,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6717778020215392},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c2ac3dcb7197d48aba4ff90d3edd397","model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5909672610445402,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5909672610445402},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"75d9086ba7d7cfc145647a7980b00931","model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.666122174975011,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.666122174975011},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfcb5e9407d6b8cdd002b83bb78f87b3","model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6459498850561176,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6459498850561176},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3850093f377c8e1e644dcdcecb95a5f1","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.547619047619048,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-astra-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.547619047619048},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"adb8a6728c00e27c92196139aaf8ba46","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.633333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-astra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.633333333333333},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"86fe2eddf752944c74455199b002138f","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-astra-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e5fe40d5d682efd309aaf5347d11f0d","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.595238095238095,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-astra-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595238095238095},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c82eb283263e7c77d91bffb51ad9d951","model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.533333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-astra-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.533333333333333},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efa59bb0628e75a7f1e32060c3eb76d4","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.414029545454545,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414029545454545},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"835213ee4a05b75c40cb1a3629634960","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.262162878787879,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262162878787879},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3a856fe2544297e8bbc46829c7407fa2","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.401351515151515,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.401351515151515},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"868fa02ea01e1cd6a6bf8e56a6501739","model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.365771717171717,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.365771717171717},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca3e95fceaf373074c8dd3cf66c6946","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.615,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-astra-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.615},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e829d764b6c12d2fb9cdd8b27d5bf1fe","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.638,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-astra:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.638},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98f3bf3f43af41a99a2c1729acb2d6c2","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.451,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-astra-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.451},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c50f8c3934df5fe5d3f37f89f1748bb","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.593,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-astra-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.593},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"41322ccaecf9f4c4b76759c33b7e83bb","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.527,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-astra-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.527},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3d45210178ea91be255ebae4a0fc02f9","model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.322,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.322},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d92ffbdc5b974e0d4e310e22baabb3ee","model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.304,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.304},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4d004744b2a8ab59763b8a0f35cc8595","model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fe8981f9b157883b75d6a024fc50a43f","model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.304,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.304},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c2b691f26fb4a04a2b468a95a940eedc","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-astra-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf87d9afdb957adc6bd1bb79bf9aa131","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.590909090909091,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-astra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.590909090909091},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97aba6fd2ba83134c68d12adf4052b2d","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.419191919191919,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-astra-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.419191919191919},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8321e0d30a8c49f67eaac2fbfaa55731","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.54040404040404,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-astra-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.54040404040404},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f69b05e3eab126d848de0c9beb9437b","model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.494949494949495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-astra-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.494949494949495},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"73ced74aaf87c5705e01727a40aa1539","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.71,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.71},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c609fd872cd481c43767a59ac5bcbe1b","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.63,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.63},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f41c9b46266c2ea1c524a3bef98524e","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.715,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.715},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13daaf52063374f48c45a72d9650cd3f","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.72,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.72},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da1b98c5169442198149be69072e532d","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1569.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1569.36},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e688c9a01c75c7b4ea7e5173dfea6abf","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1403.26,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1403.26},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c745bd1f0e53d9a2b2dc8452b7b1bcdf","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1536.04,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1536.04},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a68c9503a039e79ccae98dafa1ad65f","model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1498.47,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1498.47},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a8a6101e5554b6764b7d93818b4ac67d","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.962626262626263,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra-xhigh:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.962626262626263},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"fc992165b0b19270313820c345614fea","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.931313131313131,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra-low:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.931313131313131},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cf168ce855646f9d544a13a65c122472","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.94949494949495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra-high:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.94949494949495},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"875f23774f8f9d837b13117a3339dd0f","model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.939393939393939,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra-medium:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.939393939393939},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32ac744062378b83d1acfbec095bcd48","model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.430927835051546,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-xhigh:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra-xhigh:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.430927835051546},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"304428ba4836c1274bdcd6f4ea51f433","model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"865ba1a5ad4a26ac192c7e95fa8965a3","model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.4,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-high:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra-high:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff6cccf9d1159f946c59e43a290d88ad","model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.354639175257732,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra-medium:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra-medium:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.354639175257732}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":21,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":100},"family":"FrontierMath","row_ids":["0f175b01de79f7689d7f750dac02d89b"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"GPQA Diamond","row_ids":["c863d641570942861efdba9c19093efa"],"score":100.0},{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench-Science","row_ids":["33b37c700e39b0d0c6a88f59bf14c960"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"HLE","row_ids":["955cc9868301478204da3b4815a3ad3c"],"score":100.0},{"benchmark_version":"2 Hard","exact":{"denominator":1,"numerator":100},"family":"CritPt","row_ids":["0a974b1036bafb40411da54c5245579a"],"score":100.0}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","peer_count":39,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","peer_count":52,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705}],"score":100.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":248026779,"numerator":19458728500},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1100},"family":"tau3-Banking","row_ids":["388b55b5431573f02c7a1415569151aa"],"score":84.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5800},"family":"GDPval-AA","row_ids":["b076d23f5b3bf4c32b53727e95233455"],"score":71.6},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":275},"family":"Terminal-Bench","row_ids":["28b064c611914ab9c6c8b9c66db69c4f"],"score":91.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1725},"family":"AutomationBench","row_ids":["fb60ba99063f49f876c4c8d56120f0f3"],"score":90.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6500},"family":"AA-Briefcase","row_ids":["e3f8cbf64c84e82bb4a86c035247a89a"],"score":84.42},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":50},"family":"OSWorld","row_ids":["2ed0ffc858d7ba150a21822c243ebff1"],"score":50.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1750},"family":"AA-AnalystAgent","row_ids":["15975e3d42ab5d82a67bcbe277dd1b83"],"score":76.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","peer_count":82,"percentile":71.6,"percentile_exact":{"denominator":81,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","peer_count":49,"percentile":91.67,"percentile_exact":{"denominator":3,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":90.79,"percentile_exact":{"denominator":19,"numerator":1725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":78,"percentile":84.42,"percentile_exact":{"denominator":77,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125}],"score":78.45,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":1911,"numerator":184900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4800},"family":"FrontierCode","row_ids":["18ca4c50e781eec01bc5749fbb189315"],"score":97.96},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1200},"family":"Vals Code Migration","row_ids":["bc107d6862df5282c7cf8b968206739b"],"score":92.31},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["948d62640ac1b6f74166a175a86cf6a7"],"score":100.0}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","peer_count":50,"percentile":97.96,"percentile_exact":{"denominator":49,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":27,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":9,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1}],"score":96.76,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2954325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3675},"family":"GDP.pdf","row_ids":["96f5c4d98f214c070d53868688e0a1a3"],"score":96.71},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1650},"family":"SimpleQA Verified","row_ids":["3557cefe053ef3a2ca1c95595065295f"],"score":97.06},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3450},"family":"AA-Omniscience","row_ids":["eec9e765997481a4684eb7ff4710087d","6e1af818db0c40a97c73d45f3dd92647"],"score":65.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958}],"score":86.29,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":1377,"numerator":110350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2600},"family":"Humanity Sixth Sense","row_ids":["35c55ac70a4db5208200b7b35fe696c9"],"score":96.3},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1600},"family":"MMMU-Pro","row_ids":["ec04ac79e57d7b9ceb9bc5c32ee3fd30"],"score":94.12},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"Chartography","row_ids":["600568174d3d7470e372c6b4483b5633"],"score":50.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","peer_count":28,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0}],"score":80.14,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","peer_count":39,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","peer_count":52,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","peer_count":82,"percentile":71.6,"percentile_exact":{"denominator":81,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","peer_count":49,"percentile":91.67,"percentile_exact":{"denominator":3,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":90.79,"percentile_exact":{"denominator":19,"numerator":1725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":78,"percentile":84.42,"percentile_exact":{"denominator":77,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","peer_count":50,"percentile":97.96,"percentile_exact":{"denominator":49,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":27,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":9,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","peer_count":28,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0}],"result_updated_at":"2026-10-28","score":89.24,"score_exact":{"denominator":21283059798,"numerator":1899215940095}}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"source_id":"openai/gpt-6-astra"},"m-568e0eb3a95b5113":{"id":"m-568e0eb3a95b5113","in_window":true,"name":"Claude Opus 5","observed_records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"agents","direction":"higher_better","id":"e54e313a4cf3bafac7d6e3d2385d1025","model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":26.9,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":26.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"agents","direction":"higher_better","id":"7db5563eacfc0ff2de1106fa741f24cd","model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"ac0f80ad7501f371bea4ff3e","raw_value":52.6,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row1:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench:Terminal-Bench 4.0","source_record_id":"gpt6-astra-official:table2:row1:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":52.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"d8372588b76cfe9aa6be7425ce13cfb0","model_id":"m-568e0eb3a95b5113","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench Science 0.1; preserve page footnotes and task-specific settings","protocol_id":"6e315c8a4731951529a013b5","raw_value":30.0,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row1:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench-Science:Terminal-Bench Science 0.1","source_record_id":"gpt6-astra-official:table3:row1:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":30.0},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","model_id":"m-568e0eb3a95b5113","original_version_label":"v2","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Claude Opus 5 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"f541c2fdbde72cb43c74ebe8a432a7bf","model_id":"m-568e0eb3a95b5113","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":41.8,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:Claude Opus 5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:Claude Opus 5","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1f1de51f1003f9f9b7cc66b798316c8d","model_id":"m-568e0eb3a95b5113","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Agentic scientific research Terminal-Bench-Science 0.1³; preserve page footnotes and task-specific settings","protocol_id":"1975ca8c09f87d3fc7597c12","raw_value":29.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"29.0%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row14:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1","source_record_id":"official-0dd3bed5c7dd:table0:row14:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":29.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"Opus-5.0","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"ac8f6dd5c815b356a6b57350a5fb1c7e","model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":51.8,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"51.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column1","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":51.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"Opus-5.0","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"4fbfacefe470a5ba3c84b133b49ffefd","model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.0,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.0","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column1","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"Opus-5.0","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6fa1a31f17a7792b499d20a7e15bd3ed","model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":50.3,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"50.3","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column1","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":50.3},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"Opus-5.0","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c1b1f70a9c86ed29fb6aed62d8d0a04b","model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":84.0,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"84.0","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column1","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":84.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"Opus-5.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"d4f91d80df262b71392602c5e046be89","model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column Opus-5.0. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":93.4,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"93.4","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column1","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":93.4},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"74d5382d9b4d1ccd5a786120f5a550f0","model_id":"m-568e0eb3a95b5113","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic scientific research Terminal-Bench-Science 0.1 [1]; preserve page footnotes and task-specific settings","protocol_id":"88023c6718be09c275d82cfc","raw_value":29.0,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"29.0%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row2:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1 [1]","source_record_id":"anthropic-fable51:table0:row2:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":29.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"986d51b676eb20ae4c487a9e43e01df2","model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":52.3,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"52.3%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row4:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"anthropic-fable51:table0:row4:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":52.3},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"b2df261aa6e00c4606b5fc3f6b802f0f","model_id":"m-568e0eb3a95b5113","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":93.3,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"93.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":93.3},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","model_id":"m-568e0eb3a95b5113","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"773afe0f7c5b5cfa8f123e1e37bca353","model_id":"m-568e0eb3a95b5113","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":48.7,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"48.7","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":48.7},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"2cc0820ff84711ce0c2bfadbfbb013aa","model_id":"m-568e0eb3a95b5113","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":74.7,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"74.7","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":74.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"c5e52e0f769a492e0c855ab1119ac22b","model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5717171717171717,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1af19aa08ae13afdcc123c58"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1af19aa08ae13afdcc123c58","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5717171717171717},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"bab2666054b26f7187e9033bd8e5d708","model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5797979797979798,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eae5ec5ab864c2b64a8c4a82"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eae5ec5ab864c2b64a8c4a82","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5797979797979798},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"fab39eb2532678a78022356198cf9feb","model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38181818181818183,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3fd792d0eb662957180872d4"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"3fd792d0eb662957180872d4","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38181818181818183},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"b3d297b7a0ed2397954130a092a98e6b","model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5383838383838384,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"336d4a183af0b8f266da26be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"336d4a183af0b8f266da26be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5383838383838384},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14}],"official_evidence":{"artifact_sha256":"2f2269c3e1f9030bf9adc7e06e1979500968e3fc300eedd57a2b54424cab1694","basis":"official_release_page","date":"2026-07-24","sha256":"2f2269c3e1f9030bf9adc7e06e1979500968e3fc300eedd57a2b54424cab1694","url":"https://www.anthropic.com/news/claude-opus-5"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":3,"exact":{"denominator":2128,"numerator":133225},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":300},"family":"FrontierMath","row_ids":["88ff5cf09a8ea0336d12c5e7aac222d3"],"score":42.86},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":225},"family":"GPQA Diamond","row_ids":["e80c9d5d961020a2775a86f5a8e8c7ff"],"score":56.25},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1450},"family":"Terminal-Bench-Science","row_ids":["12c1a7b1c4eb18f0409ef269987fa5ca"],"score":76.32},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":75},"family":"HLE","row_ids":["852fa04b612ed916ac880c4c5a9b046b"],"score":75.0}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v2","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","peer_count":39,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6}],"score":62.61,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":6,"exact":{"denominator":252,"numerator":18965},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["803b3c7ebd5d583be94f718268d22790"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"GDPval-AA","row_ids":["3e2aa560fb85d043c269e4880c8d1ed7"],"score":66.67},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":325},"family":"Terminal-Bench","row_ids":["7df1cc76a5ce426695809152ab7c2b62"],"score":81.25},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"AutomationBench","row_ids":["ae96d3d1f5d89c47f0a27de7f5eca5d4"],"score":40.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":800},"family":"AA-Briefcase","row_ids":["a01ae7f3c128573c113df7e946ef1073"],"score":88.89},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":50},"family":"OSWorld","row_ids":["5f71dd3fff3f2fe7b85716b5a982127d"],"score":50.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":100},"family":"MCP Atlas","row_ids":["2ccdf17c2a9e8d9727ba57cd7a67ff5d"],"score":100.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","included":true,"model_id":"m-568e0eb3a95b5113","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","peer_count":49,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":7,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8}],"score":75.26,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":75},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":75},"family":"DeepSWE","row_ids":["d36cdba4dae45a238d9898e2ce34c241"],"score":75.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7}],"score":75.0,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":2,"exact":{"denominator":27,"numerator":1730},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1300},"family":"Humanity Sixth Sense","row_ids":["175d0f5744557d89454896c1420ebf13"],"score":48.15},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":80},"family":"Chartography","row_ids":["1ce2cb7b64555f9b41b3b5999f8e70ba"],"score":80.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","peer_count":28,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4}],"score":64.07,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v2","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","peer_count":39,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","included":true,"model_id":"m-568e0eb3a95b5113","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","peer_count":49,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":7,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","peer_count":28,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4}],"result_updated_at":"2026-09-29","score":69.84,"score_exact":{"denominator":131328,"numerator":9172165}}],"release_date":"2026-07-24","release_status":"published","release_verified":true,"source_id":"anthropic/claude-opus-5"},"m-5c18cc0e1d3391a9":{"id":"m-5c18cc0e1d3391a9","in_window":true,"name":"Mistral Medium 3.5","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526}],"official_evidence":{"artifact_sha256":"2f83ab54cf0df8335b080c25c95cc6852ada259bf85c12135f848a991e748f85","basis":"official_release_page","date":"2026-05-22","sha256":"2f83ab54cf0df8335b080c25c95cc6852ada259bf85c12135f848a991e748f85","url":"https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5/"},"primary_profile_id":"native","profiles":[{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":49583375},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1700},"family":"GPQA Diamond","row_ids":["f6a22fbde90853e88aebc327f5973f62"],"score":28.81},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["1b904acff7083099b3a91e7b6c5b575a"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":650},"family":"HLE","row_ids":["5c0267c76de42807699083ce1e0581c4"],"score":10.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":200},"family":"CritPt","row_ids":["98267d8baf13117a13191a66537ee6eb"],"score":3.92}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":28.81,"percentile_exact":{"denominator":59,"numerator":1700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","peer_count":52,"percentile":3.92,"percentile_exact":{"denominator":51,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005}],"score":15.09,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1450449,"numerator":28030885},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":350},"family":"tau3-Banking","row_ids":["d0f6a7a8e824abf26de5a2c3eb1a8991"],"score":26.92},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":200},"family":"GDPval-AA","row_ids":["641dcee193d24a80d7b9a09327dfcf69"],"score":22.22},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["47e9d9d6a84d129959dab9eed70b71bc"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"AutomationBench","row_ids":["b05e70f346089460a993097793ed0c46"],"score":25.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1300},"family":"AA-Briefcase","row_ids":["671c80eaf4d486bab2d265a1c67045fd"],"score":16.88},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"EnterpriseOps-Gym","row_ids":["da0d77fdb30d1e419d687691f9d0af7b"],"score":22.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":200},"family":"AA-AnalystAgent","row_ids":["ef6364df2840f04f7c6f211bd995980d"],"score":8.7}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","peer_count":82,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":78,"percentile":16.88,"percentile_exact":{"denominator":77,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","peer_count":24,"percentile":8.7,"percentile_exact":{"denominator":23,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125}],"score":19.33,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":1000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1000},"family":"FrontierCode","row_ids":["c2dd8ca14021780f9edc8dc0975ff8ae"],"score":20.41}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","peer_count":50,"percentile":20.41,"percentile_exact":{"denominator":49,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939}],"score":20.41,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":106,"numerator":3825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"IFBench","row_ids":["5fbfa4015dcb4ab37c1080316055d055"],"score":75.0},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"GDP.pdf","row_ids":["2bd6dbb1c5200680f554dfce1a26feb0"],"score":12.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1100},"family":"AA-Omniscience","row_ids":["2b5b8d191f3e5e3bfa651cee2bbbd819","d7d18670af809e53c710911ecd277233"],"score":20.75}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378}],"score":36.08,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":150},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":150},"family":"MMMU-Pro","row_ids":["ce0faae345d00b20396f1024d947c26c"],"score":8.82}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798}],"score":8.82,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":28.81,"percentile_exact":{"denominator":59,"numerator":1700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","peer_count":52,"percentile":3.92,"percentile_exact":{"denominator":51,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","peer_count":82,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":78,"percentile":16.88,"percentile_exact":{"denominator":77,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","peer_count":24,"percentile":8.7,"percentile_exact":{"denominator":23,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","peer_count":50,"percentile":20.41,"percentile_exact":{"denominator":49,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798}],"result_updated_at":null,"score":20.78,"score_exact":{"denominator":411223564752,"numerator":8547240422815}}],"release_date":"2026-05-22","release_status":"published","release_verified":true,"source_id":"mistral/mistral-medium-3-5"},"m-5cf6d704eda6778e":{"id":"m-5cf6d704eda6778e","in_window":true,"name":"GLM-5.2","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"official_evidence":{"artifact_sha256":"c6c90dc0946031acbba79d00549ddb918494022b454142617bed1e9492fdd97e","basis":"official_release_page","date":"2026-06-16","sha256":"c6c90dc0946031acbba79d00549ddb918494022b454142617bed1e9492fdd97e","url":"https://z.ai/blog/glm-5.2"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3,"numerator":50},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":50},"family":"GPQA Diamond","row_ids":["55c9779a030f53f08065b75a89097776"],"score":16.67}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2}],"score":16.67,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3,"numerator":55},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"APEX-Agents","row_ids":["6f3782a0a7f13ec12e5eae491c79c282"],"score":20.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["96fb24279652563e7d2bc82b71d0fc52"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":100},"family":"AA-Briefcase","row_ids":["5e2ad6a27f43522ccb21c4f280e4613e"],"score":33.33},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":20},"family":"MCP Atlas","row_ids":["d57170302e825b662bfe4acca3fbd398"],"score":20.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6}],"score":18.33,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["546ef51508bcfe588737e01c240c5015"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":17,"numerator":250},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":250},"family":"SimpleQA Verified","row_ids":["bcb4fb304c488f86e521408952f81841"],"score":14.71}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"score":14.71,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"result_updated_at":"2026-08-27","score":12.99,"score_exact":{"denominator":102,"numerator":1325}}],"release_date":"2026-06-16","release_status":"published","release_verified":true,"source_id":"zhipuai/glm-5.2"},"m-5ea285ac7903cbc7":{"id":"m-5ea285ac7903cbc7","in_window":true,"name":"Kimi K2.7 Code","observed_records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1239,"numerator":75700},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4400},"family":"GPQA Diamond","row_ids":["7516363fe155a8bbdfc789cee4300773"],"score":74.58},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1000},"family":"HLE","row_ids":["7c461dbccbd2115e299061bd3cfb0f10"],"score":47.62}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":74.58,"percentile_exact":{"denominator":59,"numerator":4400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","peer_count":64,"percentile":47.62,"percentile_exact":{"denominator":21,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148}],"score":61.1,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":440154,"numerator":17585753},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":575},"family":"tau3-Banking","row_ids":["1b7ece54bb0355e69302fd5373585994"],"score":44.23},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1000},"family":"GDPval-AA","row_ids":["ae60755488af790d34ab9ed3aa726374"],"score":37.04},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["3126ca26ac0a6c7dab9a15033082575d"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":850},"family":"AutomationBench","row_ids":["2f9ec0e2d5074f31fb106a7f0db810e6"],"score":44.74},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":300},"family":"AA-Briefcase","row_ids":["8c6b16e6477c30b02c0cf413d3d98aff"],"score":27.27},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"EnterpriseOps-Gym","row_ids":["1d959564327d6d5496bf9174d774a4c8"],"score":44.44}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","peer_count":82,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","peer_count":78,"percentile":27.27,"percentile_exact":{"denominator":11,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996}],"score":39.95,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":2,"exact":{"denominator":3876,"numerator":157525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":225},"family":"IFBench","row_ids":["15ec6c548623aedd9b87ffa73167943d"],"score":56.25},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["60b6aea7e19a90a921731bc7d92e251c"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":300},"family":"SimpleQA Verified","row_ids":["29624d33497b02671ed8ec8a79e7f2e8"],"score":17.65}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365}],"score":40.64,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":74.58,"percentile_exact":{"denominator":59,"numerator":4400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","peer_count":64,"percentile":47.62,"percentile_exact":{"denominator":21,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","peer_count":82,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","peer_count":78,"percentile":27.27,"percentile_exact":{"denominator":11,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365}],"result_updated_at":"2026-08-27","score":47.7,"score_exact":{"denominator":43264497276,"numerator":2063781354715}}],"release_date":"2026-06-12","release_status":"published","release_verified":true,"source_id":"moonshotai/kimi-k2.7-code"},"m-60cd64beda366768":{"id":"m-60cd64beda366768","in_window":true,"name":"MiMo-V2.6-Flash","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","model_id":"m-60cd64beda366768","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","model_id":"m-60cd64beda366768","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f38424924d0f84fbc9a6a159bb2bb1d4","model_id":"m-60cd64beda366768","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","model_id":"m-60cd64beda366768","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","model_id":"m-60cd64beda366768","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","model_id":"m-60cd64beda366768","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","model_id":"m-60cd64beda366768","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","model_id":"m-60cd64beda366768","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","model_id":"m-60cd64beda366768","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","model_id":"m-60cd64beda366768","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ed311eb703cfe1b731155e091addc7e","model_id":"m-60cd64beda366768","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.227272727272727,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mimo-v2-6-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.227272727272727},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","model_id":"m-60cd64beda366768","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56}],"official_evidence":{"artifact_sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","basis":"official_release_page","date":"2026-09-22","sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","url":"https://mimo.mi.com/docs/en-US/updates/model"},"primary_profile_id":"native","profiles":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":13923,"numerator":703600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["e9c81f27a54c74df81f07409ddb33f17"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":300},"family":"HLE","row_ids":["433cbb1c2fcf649c36b004a451515cf8"],"score":42.86},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["737809004b185d439bc3bdb639d99a34"],"score":54.9}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","included":true,"model_id":"m-60cd64beda366768","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","included":true,"model_id":"m-60cd64beda366768","original_version_label":"Diamond","peer_count":64,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":50.54,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":1896048,"numerator":133400075},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6100},"family":"GDPval-AA","row_ids":["42544e43967725b8e648b9b8615120d6"],"score":75.31},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":575},"family":"Terminal-Bench","row_ids":["3fbf327bd9cbc7d6e3195a9d909b9231"],"score":47.92},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1550},"family":"AutomationBench","row_ids":["250157187c56354b4eed518f47426340"],"score":81.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5900},"family":"AA-Briefcase","row_ids":["e5699451024628001ea794ad1191793b"],"score":76.62}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2.1","peer_count":82,"percentile":75.31,"percentile_exact":{"denominator":81,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","included":true,"model_id":"m-60cd64beda366768","original_version_label":"4.0","peer_count":49,"percentile":47.92,"percentile_exact":{"denominator":12,"numerator":575},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","included":true,"model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":78,"percentile":76.62,"percentile_exact":{"denominator":77,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83}],"score":70.36,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":27425},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2900},"family":"FrontierCode","row_ids":["d9ee15576eab5d8aeca7fb1035d14154"],"score":59.18},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":350},"family":"Vals Code Migration","row_ids":["501724f84396d924c42fcaca484ef573"],"score":26.92}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","included":true,"model_id":"m-60cd64beda366768","original_version_label":"2.0","peer_count":50,"percentile":59.18,"percentile_exact":{"denominator":49,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","included":true,"model_id":"m-60cd64beda366768","original_version_label":"1.1","peer_count":27,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55}],"score":43.05,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":36975},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":750},"family":"GDP.pdf","row_ids":["0f9316648f89e3d521fa219b6361fd6c"],"score":39.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["084eb35df8b2a6c0ecbd814a3709a769","b4e27389bf022ba2e7fea370cfc23e71"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253}],"score":36.72,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":10000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":700},"family":"Humanity Sixth Sense","row_ids":["672a5ff05bade6e76dd2270c17366a35"],"score":25.93},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":300},"family":"MMMU-Pro","row_ids":["f2f6f962e776bf31934ea15f004e8009"],"score":17.65}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current public protocol","peer_count":28,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289}],"score":21.79,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","included":true,"model_id":"m-60cd64beda366768","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","included":true,"model_id":"m-60cd64beda366768","original_version_label":"Diamond","peer_count":64,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2.1","peer_count":82,"percentile":75.31,"percentile_exact":{"denominator":81,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","included":true,"model_id":"m-60cd64beda366768","original_version_label":"4.0","peer_count":49,"percentile":47.92,"percentile_exact":{"denominator":12,"numerator":575},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","included":true,"model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":78,"percentile":76.62,"percentile_exact":{"denominator":77,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","included":true,"model_id":"m-60cd64beda366768","original_version_label":"2.0","peer_count":50,"percentile":59.18,"percentile_exact":{"denominator":49,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","included":true,"model_id":"m-60cd64beda366768","original_version_label":"1.1","peer_count":27,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current public protocol","peer_count":28,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289}],"result_updated_at":null,"score":48.36,"score_exact":{"denominator":621835486272,"numerator":30069425541005}}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"source_id":"xiaomi/mimo-v2.6-flash"},"m-65378fc44ea2de26":{"id":"m-65378fc44ea2de26","in_window":true,"name":"DeepSeek V4 Flash Vision","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","model_id":"m-65378fc44ea2de26","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":3717,"numerator":278450},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5000},"family":"GPQA Diamond","row_ids":["185f07372e1def6a3525e6b1b6c635f5"],"score":84.75},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4100},"family":"HLE","row_ids":["c42c9d1f70b755ec31ff5dcd6c984f87"],"score":65.08}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":84.75,"percentile_exact":{"denominator":59,"numerator":5000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","peer_count":64,"percentile":65.08,"percentile_exact":{"denominator":63,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211}],"score":74.91,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":105856906},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1075},"family":"tau3-Banking","row_ids":["b1886f41be6be37742b893448f0c7326"],"score":82.69},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1900},"family":"GDPval-AA","row_ids":["6a2341b888c3c4634de9d8d25fc69ba9"],"score":70.37},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":172},"family":"Terminal-Bench","row_ids":["cdfb425c1664f272548a144cf79359c8"],"score":57.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1075},"family":"AutomationBench","row_ids":["6ca0b1280bb8cb8f57834c52f2472139"],"score":56.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5200},"family":"AA-Briefcase","row_ids":["94b4404155c32f58241674a265734f95"],"score":67.53},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":700},"family":"EnterpriseOps-Gym","row_ids":["54f26e677fb3a678985d92f416cf28ad"],"score":77.78}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","peer_count":82,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"4.0","peer_count":76,"percentile":57.33,"percentile_exact":{"denominator":3,"numerator":172},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","peer_count":78,"percentile":67.53,"percentile_exact":{"denominator":77,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943}],"score":68.71,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":1050},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1050},"family":"GDP.pdf","row_ids":["b7c40d3349d6abfac4e0b93abf2adfa2"],"score":55.26}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12}],"score":55.26,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":500},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":500},"family":"MMMU-Pro","row_ids":["dbca8fa745f042a216711b700c5e06ca"],"score":29.41}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717}],"score":29.41,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":84.75,"percentile_exact":{"denominator":59,"numerator":5000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","peer_count":64,"percentile":65.08,"percentile_exact":{"denominator":63,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","peer_count":82,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"4.0","peer_count":76,"percentile":57.33,"percentile_exact":{"denominator":3,"numerator":172},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","peer_count":78,"percentile":67.53,"percentile_exact":{"denominator":77,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717}],"result_updated_at":null,"score":62.38,"score_exact":{"denominator":3090321234,"numerator":192760715305}}],"release_date":"2026-08-21","release_status":"published","release_verified":true,"source_id":"published/aa:deepseek-v4-flash-vision"},"m-65a8fc1dec61133f":{"id":"m-65a8fc1dec61133f","in_window":true,"name":"GPT-5.6 Luna","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"bc79c1e1910e14e0ff6ab222415c13af","model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":79.5,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","model_id":"m-65a8fc1dec61133f","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"ee6e51ce0e1c80e8c9e718c1bdb9fdb9","model_id":"m-65a8fc1dec61133f","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":58.5,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":58.5},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"GPT-5.6 Luna","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7ef9b74eae323671beacf43bf25de9bf","model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":67.0,"research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"67%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column5"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column5","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":67.0},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"9731cf222852cec61d24dbd978909d67","model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.13079299691040164,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a1872416ae701ca8fb4e8991"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"a1872416ae701ca8fb4e8991","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.13079299691040164},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"fc9f1ec47101e426fc54439fc6b50272","model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.20202020202020202,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"542d10b151a6029af2b51f8e"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"542d10b151a6029af2b51f8e","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.20202020202020202},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"998e09d4ab7b7e92640376de5c61a4b7","model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3484848484848485,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3ae87974f36f55900ac2e65c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"3ae87974f36f55900ac2e65c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3484848484848485},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"28bd42f626108b631bb1a28efa0c3dd4","model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.298989898989899,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0ace10d1f7f2b98ac55a15f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"f0ace10d1f7f2b98ac55a15f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.298989898989899},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":625},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":0},"family":"FrontierMath","row_ids":["0c9a231c5ad369432b2d4e6968b6bc7c"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":25},"family":"GPQA Diamond","row_ids":["710e7d489e0ed28e0e46334dd5f83f89"],"score":12.5},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["85fbb43a0855c52959cd8ec4413297f0"],"score":3.95}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v2","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"score":5.48,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":54,"numerator":1475},"families":[{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":25},"family":"Terminal-Bench","row_ids":["5bec3b19db8e7a3362bcd865766b2cc8"],"score":25.0},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"AutomationBench","row_ids":["51dfce6bb3b25136b0ea3545031f03b8"],"score":12.5},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":400},"family":"AA-Briefcase","row_ids":["06516697c2ae14f64607af00c620c8a8"],"score":44.44}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","peer_count":49,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798}],"score":27.31,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":25},"family":"DeepSWE","row_ids":["b632e7c137b19cc34931e9481ad27bf9"],"score":12.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2}],"score":12.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"GDP.pdf","row_ids":["706d4a87bdc7adf93c73d7e4099021c8"],"score":25.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7}],"score":25.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["6698824b06580d547cc03c11c24e9e05"],"score":3.7},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["b27c843f9cde1b388c12022729f0ceb0"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","peer_count":28,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"reference_revision":"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4}],"score":1.85,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v2","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","peer_count":49,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","peer_count":28,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"reference_revision":"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4}],"result_updated_at":"2026-07-09","score":15.88,"score_exact":{"denominator":684,"numerator":10865}}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"source_id":"openai/gpt-5.6-luna"},"m-66c1aae520b7679a":{"id":"m-66c1aae520b7679a","in_window":true,"name":"Kimi K3.1 Pro","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3.1 Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"952e1d48294b90de64278faf65544562","model_id":"m-66c1aae520b7679a","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.64,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.64},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3.1 Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"58dd7cae94574219da61327aad86af6e","model_id":"m-66c1aae520b7679a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.43,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"31.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row8","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":31.43},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","model_id":"m-66c1aae520b7679a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","model_id":"m-66c1aae520b7679a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","model_id":"m-66c1aae520b7679a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"official_evidence":{"artifact_sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","basis":"official_release_page","date":"2026-10-24","sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","url":"https://www.kimi.com/en/blog/"},"primary_profile_id":"native","profiles":[{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":41769,"numerator":3441925},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1075},"family":"Terminal-Bench-Science","row_ids":["440fdb6baf103b127060630dee151d5e"],"score":82.69},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5300},"family":"HLE","row_ids":["8cb18c97abf2e38058caf77fed29461c"],"score":84.13},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4100},"family":"CritPt","row_ids":["e4f3938369b9a55c0c1841556cc18184"],"score":80.39}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"0.1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","peer_count":64,"percentile":84.13,"percentile_exact":{"denominator":63,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","peer_count":52,"percentile":80.39,"percentile_exact":{"denominator":51,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495}],"score":82.4,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2725569,"numerator":238699840},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":8000},"family":"GDPval-AA","row_ids":["948bda542167bd107b882e3640f83a0c"],"score":98.77},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench","row_ids":["1c21aa90c01783fcd785c617ba4d90f9"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1850},"family":"AutomationBench","row_ids":["97fcee40bdfab50663d8d9aa18bd9131"],"score":97.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7400},"family":"AA-Briefcase","row_ids":["2b512abdb78eff5062c8cbfe38f0a197"],"score":96.1},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1050},"family":"AA-AnalystAgent","row_ids":["8b93841abeaaa5b29864bb557e6a9da7"],"score":45.65}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","peer_count":82,"percentile":98.77,"percentile_exact":{"denominator":81,"numerator":8000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"4.0","peer_count":76,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":78,"percentile":96.1,"percentile_exact":{"denominator":77,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"score":87.58,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":7,"numerator":500},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":500},"family":"FrontierCode","row_ids":["2cd7a142e0f06d143bb474c89fce8084"],"score":71.43}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"2.0","peer_count":50,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444}],"score":71.43,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":165875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1875},"family":"GDP.pdf","row_ids":["24d56f9893e1402bb483f6bef166ba08"],"score":98.68},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3500},"family":"AA-Omniscience","row_ids":["e4c0067c2ce56fc8c2df02c769685194","faf92c6808beb316ccdbea4a34863bce"],"score":66.04}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008}],"score":82.36,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1400},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1400},"family":"MMMU-Pro","row_ids":["75435cf97a5f0d98b110b2544004171f"],"score":82.35}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422}],"score":82.35,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"0.1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","peer_count":64,"percentile":84.13,"percentile_exact":{"denominator":63,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","peer_count":52,"percentile":80.39,"percentile_exact":{"denominator":51,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","peer_count":82,"percentile":98.77,"percentile_exact":{"denominator":81,"numerator":8000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"4.0","peer_count":76,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":78,"percentile":96.1,"percentile_exact":{"denominator":77,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"2.0","peer_count":50,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422}],"result_updated_at":null,"score":81.49,"score_exact":{"denominator":127698358788,"numerator":10405967280655}}],"release_date":"2026-10-24","release_status":"published","release_verified":true,"source_id":"moonshotai/kimi-k3.1-pro"},"m-68480fc0423c5034":{"id":"m-68480fc0423c5034","in_window":true,"name":"Muse Wonder","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","model_id":"m-68480fc0423c5034","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"cbbd47b78fc1abe8ff45d7f2b54da87d","model_id":"m-68480fc0423c5034","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":49.49,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"49.49%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":49.49},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Wonder / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"adff8e814a52d75cbf97b5acfa5177a4","model_id":"m-68480fc0423c5034","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"22.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row11","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":22.86},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","model_id":"m-68480fc0423c5034","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e115891c0c637d1c56c96850db20ff4","model_id":"m-68480fc0423c5034","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1609.93,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1609.93},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","model_id":"m-68480fc0423c5034","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2644e5f192597eea190a353c9183fee4","model_id":"m-68480fc0423c5034","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.382,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.382},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","model_id":"m-68480fc0423c5034","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9eae38d8dd77fe54bea973dd6df61068","model_id":"m-68480fc0423c5034","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7783171521035599,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7783171521035599},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","model_id":"m-68480fc0423c5034","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9811cade034a990d6c7548b60ff78c2","model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6722179886070812,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6722179886070812},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"905e3245f89cdbd58e66ffa998c7d55b","model_id":"m-68480fc0423c5034","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.219047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219047619047619},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","model_id":"m-68480fc0423c5034","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf9770ca0af5aaab8bd90674be0e9c44","model_id":"m-68480fc0423c5034","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.266252272727273,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.266252272727273},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","model_id":"m-68480fc0423c5034","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3bdbdb2f3d97db0aeabbd6b686df8881","model_id":"m-68480fc0423c5034","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.345,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.345},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c7cb147aa61e00d80a403c5843ea98f0","model_id":"m-68480fc0423c5034","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.39,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c38f3af66f227220bc59dfa2fcaefe92","model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.292,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"466d76101a1f658f491b4db06c6a2528","model_id":"m-68480fc0423c5034","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.383838383838384,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.383838383838384},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","model_id":"m-68480fc0423c5034","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b04f85cac6b04b529f5d4a3a6c7488f8","model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.51,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.51},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","model_id":"m-68480fc0423c5034","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":15912,"numerator":1118525},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":500},"family":"FrontierMath","row_ids":["a21e66eb62f2bfca72e7c625965b008a"],"score":55.56},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1050},"family":"Terminal-Bench-Science","row_ids":["161dc1424fa043da938a084c4e822aa1"],"score":80.77},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":125},"family":"HLE","row_ids":["c73228fc930f557212149ca3c921d3fb"],"score":62.5},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1400},"family":"CritPt","row_ids":["024c0dcfccf03a6bb518b6c682477674"],"score":82.35}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"0.1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Diamond","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","peer_count":52,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505}],"score":70.29,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":21649031},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7700},"family":"GDPval-AA","row_ids":["5449b5fbd439528611d565a71ad28a45"],"score":95.06},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":262},"family":"Terminal-Bench","row_ids":["6fe9193a0e51ae328fa8f19ad476b5c1"],"score":87.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1700},"family":"AutomationBench","row_ids":["b2ddde3d390d97ad479920a7865821c3"],"score":89.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7200},"family":"AA-Briefcase","row_ids":["849c220b415c509e507a4751daf48b0a"],"score":93.51}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2.1","peer_count":82,"percentile":95.06,"percentile_exact":{"denominator":81,"numerator":7700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":78,"percentile":93.51,"percentile_exact":{"denominator":77,"numerator":7200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65}],"score":91.34,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":51000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3700},"family":"FrontierCode","row_ids":["3733f77f03c95d14f61b1419adf05cd6"],"score":75.51},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1100},"family":"Vals Code Migration","row_ids":["ab784b677fcf21bc1b248358effdf278"],"score":84.62}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"2.0","peer_count":50,"percentile":75.51,"percentile_exact":{"denominator":49,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"1.1","peer_count":27,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36}],"score":80.06,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":85700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1800},"family":"GDP.pdf","row_ids":["a4934e52e2688102b1f21d91ca098935"],"score":94.74},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4000},"family":"AA-Omniscience","row_ids":["3f07a1a0073c74e060ddc96a3f9e8bcb","58f09829d0f39ffd8664ffd449b9bfbd"],"score":75.47}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837}],"score":85.1,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":37775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2300},"family":"Humanity Sixth Sense","row_ids":["b9ee5aa138e25de06492fbd597ad313a"],"score":85.19},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1350},"family":"MMMU-Pro","row_ids":["bc79e5409c84a79a9897609e9d7ac22e"],"score":79.41}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","peer_count":28,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","peer_count":35,"percentile":79.41,"percentile_exact":{"denominator":17,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341}],"score":82.3,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"0.1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Diamond","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","peer_count":52,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2.1","peer_count":82,"percentile":95.06,"percentile_exact":{"denominator":81,"numerator":7700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":78,"percentile":93.51,"percentile_exact":{"denominator":77,"numerator":7200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"2.0","peer_count":50,"percentile":75.51,"percentile_exact":{"denominator":49,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"1.1","peer_count":27,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","peer_count":28,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","peer_count":35,"percentile":79.41,"percentile_exact":{"denominator":17,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341}],"result_updated_at":"2026-10-12","score":81.67,"score_exact":{"denominator":310917743136,"numerator":25393529124989}}],"release_date":"2026-10-12","release_status":"published","release_verified":true,"source_id":"meta/muse-wonder-1.0"},"m-6ac02cf910375655":{"id":"m-6ac02cf910375655","in_window":true,"name":"LongCat 2.0","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","model_id":"m-6ac02cf910375655","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","model_id":"m-6ac02cf910375655","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","model_id":"m-6ac02cf910375655","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","model_id":"m-6ac02cf910375655","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-29","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":177,"numerator":6250},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2200},"family":"GPQA Diamond","row_ids":["7c995d56b5fa39eef927afa2cc879ab9"],"score":37.29},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":100},"family":"HLE","row_ids":["12e6c7bd6a0e1d4d1047be66dc2b360a"],"score":33.33}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":37.29,"percentile_exact":{"denominator":59,"numerator":2200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond","peer_count":64,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115}],"score":35.31,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":32683129},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":275},"family":"tau3-Banking","row_ids":["69ddb67fade6ab2b7cbae1dad6474f61"],"score":21.15},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2200},"family":"GDPval-AA","row_ids":["d1bad78363493513c436f20f24fa1380"],"score":27.16},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["27a6f0107d1b80cc1605307626b6cc35"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":400},"family":"AutomationBench","row_ids":["5cc05f32d32a7a675fc47d1b76ed156a"],"score":21.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1800},"family":"AA-Briefcase","row_ids":["1e5075fa69ac15167616fbc24b2c2ddf"],"score":23.38}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":21.15,"percentile_exact":{"denominator":13,"numerator":275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v2.1","peer_count":82,"percentile":27.16,"percentile_exact":{"denominator":81,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":21.05,"percentile_exact":{"denominator":19,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v1.1","peer_count":78,"percentile":23.38,"percentile_exact":{"denominator":77,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73}],"score":21.22,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1175},"family":"GDP.pdf","row_ids":["c1a1527152276d41667adbe5c22b811f"],"score":30.92}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"score":30.92,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":37.29,"percentile_exact":{"denominator":59,"numerator":2200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond","peer_count":64,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":21.15,"percentile_exact":{"denominator":13,"numerator":275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v2.1","peer_count":82,"percentile":27.16,"percentile_exact":{"denominator":81,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":21.05,"percentile_exact":{"denominator":19,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v1.1","peer_count":78,"percentile":23.38,"percentile_exact":{"denominator":77,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"result_updated_at":null,"score":29.02,"score_exact":{"denominator":1272485214,"numerator":36930684955}}],"release_date":"2026-06-29","release_status":"published","release_verified":true,"source_id":"meituan/longcat-2.0"},"m-6b5176ed51518e35":{"id":"m-6b5176ed51518e35","in_window":true,"name":"Ling 3.0 Tiny","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","model_id":"m-6b5176ed51518e35","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":1500},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1500},"family":"GPQA Diamond","row_ids":["3e2102f0b8440ab308d1e2fedeb7c1ef"],"score":25.42}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":25.42,"percentile_exact":{"denominator":59,"numerator":1500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434}],"score":25.42,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3081078,"numerator":55952933},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":1225},"family":"tau3-Banking","row_ids":["e330721d27c5f1f48579252fe1951f5e"],"score":47.12},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1300},"family":"GDPval-AA","row_ids":["f9e9baed830cb24bcff4431e5893ddf5"],"score":16.05},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["9fef6d5b9c29a36ef4ec13143b2d54ae"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":25},"family":"AutomationBench","row_ids":["a389be9490eced3c5e8431f3790a6ecd"],"score":1.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1000},"family":"AA-Briefcase","row_ids":["662e05367e55e07efb079daf8cc94493"],"score":12.99}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","peer_count":82,"percentile":16.05,"percentile_exact":{"denominator":81,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":78,"percentile":12.99,"percentile_exact":{"denominator":77,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91}],"score":18.16,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":75025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":125},"family":"GDP.pdf","row_ids":["7738f3afb94846ed2b13073b7d2cb3b1"],"score":3.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["97a4da3275fb19d0e062184a2de4a918","ff2df3ac7046d17ae8e64e925f03f746"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413}],"score":18.63,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":25.42,"percentile_exact":{"denominator":59,"numerator":1500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","peer_count":82,"percentile":16.05,"percentile_exact":{"denominator":81,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":78,"percentile":12.99,"percentile_exact":{"denominator":77,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413}],"result_updated_at":null,"score":20.89,"score_exact":{"denominator":134883432684,"numerator":2817358651405}}],"release_date":"2026-08-06","release_status":"published","release_verified":true,"source_id":"published/aa:ling-3-0-tiny"},"m-71070d6294f8dc57":{"id":"m-71070d6294f8dc57","in_window":true,"name":"Quasar 438B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","model_id":"m-71070d6294f8dc57","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":117650},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1300},"family":"GPQA Diamond","row_ids":["b5ca8a579b13fb2adc96c2c993971f7c"],"score":22.03},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2600},"family":"HLE","row_ids":["92b0eaaad67a910391a173c1c1816701"],"score":41.27}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":22.03,"percentile_exact":{"denominator":59,"numerator":1300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","peer_count":64,"percentile":41.27,"percentile_exact":{"denominator":63,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131}],"score":31.65,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":391169138},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":825},"family":"tau3-Banking","row_ids":["9ed3eb345da97e474521a1cd387dd7bc"],"score":63.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4300},"family":"GDPval-AA","row_ids":["95847d736743738d4e269c199a41aca9"],"score":53.09},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["84fdb274483242a5de60a1387cb766c1"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":825},"family":"AutomationBench","row_ids":["94674bbf056258753d15a53bade2e692"],"score":43.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4000},"family":"AA-Briefcase","row_ids":["f0265d9e7d722340c4dcae3d534155b9"],"score":51.95}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":63.46,"percentile_exact":{"denominator":13,"numerator":825},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","peer_count":82,"percentile":53.09,"percentile_exact":{"denominator":81,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":43.42,"percentile_exact":{"denominator":19,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","peer_count":78,"percentile":51.95,"percentile_exact":{"denominator":77,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55}],"score":50.78,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1075},"family":"GDP.pdf","row_ids":["3fef1cbee3ec381cc5e40ecf79f229ab"],"score":28.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":28.29,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":22.03,"percentile_exact":{"denominator":59,"numerator":1300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","peer_count":64,"percentile":41.27,"percentile_exact":{"denominator":63,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":63.46,"percentile_exact":{"denominator":13,"numerator":825},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","peer_count":82,"percentile":53.09,"percentile_exact":{"denominator":81,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":43.42,"percentile_exact":{"denominator":19,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","peer_count":78,"percentile":51.95,"percentile_exact":{"denominator":77,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"result_updated_at":null,"score":37.52,"score_exact":{"denominator":636242607,"numerator":23874290621}}],"release_date":"2026-08-10","release_status":"published","release_verified":true,"source_id":"published/aa:quasar-438b"},"m-7187cf2b01d41935":{"id":"m-7187cf2b01d41935","in_window":true,"name":"Grok 4.7","observed_records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"grok-4.7","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"da019b1bad73b978b878b2436172748d","model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":23.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"633f0faf32b756ea9605dfaf"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"633f0faf32b756ea9605dfaf","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":23.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bcf918003f716a22bcd694aa5c29a3d6","model_id":"m-7187cf2b01d41935","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":28.79,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"28.79%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":28.79},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","model_id":"m-7187cf2b01d41935","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"08948c2763d27f23d79aa95e850a2638","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1493.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1493.09},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9921e04afdee4f0e2c8bdaf78946617b","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1629.15,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1629.15},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ee1cae45e4daef2fc49a871c2d21bf2","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4573333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4573333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e3860fb2dc8fd8133af52f9095ffa5bc","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4176666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4176666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"519bde5464817ff9ef16ed4852b1f40e","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6621621621621622,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6621621621621622},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fa0aba8fe53214db1bcea627dd15d8e7","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7979393245563824,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7979393245563824},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0c7f4f4f2966e204f230f128065780bd","model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5673837407466283,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673837407466283},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c5c034abc9b1f9258d1f8e6ed61c15b3","model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6351937297314658,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6351937297314658},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"71fe042f1876dfd3dd85559bcf36b99e","model_id":"m-7187cf2b01d41935","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"grok-4-7-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fea9e2aa6ef97cf3905b940cd9bbe27","model_id":"m-7187cf2b01d41935","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0952380952380952,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"grok-4-7-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0952380952380952},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"682e14ea89cac38d7b78eec67be7db59","model_id":"m-7187cf2b01d41935","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.114285714285714,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"grok-4-7:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114285714285714},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2202bb542930f8cb07391e8c21232dcf","model_id":"m-7187cf2b01d41935","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.218224494949495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218224494949495},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"960de50ed3d067e2432fea3ec13a7dec","model_id":"m-7187cf2b01d41935","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.277836363636364,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.277836363636364},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","model_id":"m-7187cf2b01d41935","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"fb4e3eeda82eaad43d73390d4436b34d","model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.185,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ecb1aa0777cad6456788f72ef77415e7","model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.236,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.236},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3ad9495b98d112315d4ec11f093491b0","model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.214,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.214},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c8bf65d630d9c75555cdbeb9072fae6","model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.232,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af304dee21e01f415a7d570dc4284b75","model_id":"m-7187cf2b01d41935","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.161616161616162,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.161616161616162},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"521b9ed2dbcfa39c75783b3521c60c17","model_id":"m-7187cf2b01d41935","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","model_id":"m-7187cf2b01d41935","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"81e891a07c20e94e9954c27d6dabbb18","model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ab15fccc9fa7f5a9f5d6e593abd0427","model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.37,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d00ca7cc7ca6c5083747c9b27e80d923","model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1591.4,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1591.4},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7139582db7642e4a86fa949b6bb86e37","model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1715.8,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1715.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6abb453800f36b2b2a1222987fa214a7","model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.55,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-high:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7-high:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.55},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":61047,"numerator":3863750},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["5f766555b014c71ff2d651ccb2fc76d7"],"score":57.89},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4300},"family":"HLE","row_ids":["353c470ec7dd55d7fbbc616285180c31"],"score":68.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3250},"family":"CritPt","row_ids":["539eef88fa72b976d41f8f0e6b20d640"],"score":63.73}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","peer_count":64,"percentile":68.25,"percentile_exact":{"denominator":63,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335}],"score":63.29,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2725569,"numerator":228484706},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7300},"family":"GDPval-AA","row_ids":["4d4ab3fcd8c470a4ca605a1e507ade3e"],"score":90.12},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":70},"family":"Terminal-Bench","row_ids":["a073820ddc97f09d7966e361779cf68e"],"score":70.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1600},"family":"AutomationBench","row_ids":["5dc195401c512a05556701293da42306"],"score":84.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7100},"family":"AA-Briefcase","row_ids":["7ce9f4dd8dbbce093592c7a8603e6868"],"score":92.21},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1900},"family":"AA-AnalystAgent","row_ids":["1c70059543aa3443966891f5d87280d7"],"score":82.61}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","peer_count":82,"percentile":90.12,"percentile_exact":{"denominator":81,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":84.21,"percentile_exact":{"denominator":19,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":78,"percentile":92.21,"percentile_exact":{"denominator":77,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","peer_count":24,"percentile":82.61,"percentile_exact":{"denominator":23,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":83.83,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":43725},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3900},"family":"FrontierCode","row_ids":["efab926973de754a7bc887e6a5732e79"],"score":79.59},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":750},"family":"Vals Code Migration","row_ids":["575f9b1e7b6ccadadc3e13c6241b21d4"],"score":57.69}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"2.0","peer_count":50,"percentile":79.59,"percentile_exact":{"denominator":49,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"1.1","peer_count":27,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22}],"score":68.64,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":901,"numerator":65725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"GDP.pdf","row_ids":["06f207d047d6194d347719cbe681b11e"],"score":75.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1050},"family":"SimpleQA Verified","row_ids":["1bbbdc3f7be54b8c3de3afd5eb04f4c5"],"score":61.76},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4350},"family":"AA-Omniscience","row_ids":["86bdedd227d21984b4d30c6ee2bdc8b4","5789f26e6b16a0b75b183dcbb8cf44bd"],"score":82.08}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947}],"score":72.95,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":700},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":700},"family":"MMMU-Pro","row_ids":["6a15ad1c8de4bbfb578c4f789e53c9b8"],"score":41.18}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341}],"score":41.18,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","peer_count":64,"percentile":68.25,"percentile_exact":{"denominator":63,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","peer_count":82,"percentile":90.12,"percentile_exact":{"denominator":81,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":84.21,"percentile_exact":{"denominator":19,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":78,"percentile":92.21,"percentile_exact":{"denominator":77,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","peer_count":24,"percentile":82.61,"percentile_exact":{"denominator":23,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"2.0","peer_count":50,"percentile":79.59,"percentile_exact":{"denominator":49,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"1.1","peer_count":27,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341}],"result_updated_at":"2026-09-22","score":69.22,"score_exact":{"denominator":223472127879,"numerator":15467797601309}}],"release_date":"2026-09-21","release_status":"published","release_verified":true,"source_id":"xai/grok-4.7"},"m-7190ddc26a5a37a0":{"id":"m-7190ddc26a5a37a0","in_window":true,"name":"Qwen3.7 Plus","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"9c0623cc90856dd0639a89861f73ff03","model_id":"m-7190ddc26a5a37a0","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen37-plus; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"79.1","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row36:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen37-plus:table0:row36:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":79.1},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"4bee782a627e367cbb2995c9ffc85564","model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":90.3,"research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"90.3","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row24:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen37-plus:table0:row24:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":90.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"e6e7f5a00f3abe3c013f4beacb96537c","model_id":"m-7190ddc26a5a37a0","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":79.0,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.0","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"e1daf1b784cda878658887e206aebf2f","model_id":"m-7190ddc26a5a37a0","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.1,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.1","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":531,"numerator":32500},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4600},"family":"GPQA Diamond","row_ids":["b8284820623630cb7a5150e367b7bed4"],"score":77.97},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":400},"family":"HLE","row_ids":["00361f2a88ddce87fbff29f4d32f2681"],"score":44.44}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":77.97,"percentile_exact":{"denominator":59,"numerator":4600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","peer_count":64,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139}],"score":61.21,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":13864851,"numerator":337939913},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["274b775cdefb96d48e1e4a5a4e4f52de"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["9202014a7bc9d24df89fe2534ffc35bc"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1900},"family":"GDPval-AA","row_ids":["c77b3a8d70f27e7507bc297339631f46"],"score":23.46},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["a84f629da93a24a910222138960c0c05"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":675},"family":"AutomationBench","row_ids":["1915dcb5be7d210d57854519bafaa4ef"],"score":35.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2600},"family":"AA-Briefcase","row_ids":["397f726491d251a65df1f043233be5de"],"score":33.77},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["e323a2908ddfad0f233fccef67ea0b29"],"score":50.0},{"benchmark_version":"4","exact":{"denominator":1,"numerator":0},"family":"BFCL","row_ids":["355d178390cc7668a17845fc6539c179"],"score":0.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["a05fb15039bcbd3145b6e77f10451f52"],"score":0.0}],"family_count":9,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","peer_count":82,"percentile":23.46,"percentile_exact":{"denominator":81,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","peer_count":78,"percentile":33.77,"percentile_exact":{"denominator":77,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2}],"score":24.37,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["9bf5f2dc87f09a1babb07763838a2e08"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":152,"numerator":9275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":175},"family":"IFBench","row_ids":["c4485d942b9b21c7a152b1888b11b7c4"],"score":87.5},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1075},"family":"GDP.pdf","row_ids":["850da6d9475999744f930f21d7861dd9"],"score":56.58},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"IFEval","row_ids":["5ec142fbe451bd07f915ecd120f2faec"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"MMLU-Pro","row_ids":["7be3425578af009ce953f32bc09ba9aa"],"score":0.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5}],"score":61.02,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1100},"family":"MMMU-Pro","row_ids":["b6422caa44d095e13cde40b5c1a62bc3"],"score":64.71}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647}],"score":64.71,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":77.97,"percentile_exact":{"denominator":59,"numerator":4600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","peer_count":64,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","peer_count":82,"percentile":23.46,"percentile_exact":{"denominator":81,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","peer_count":78,"percentile":33.77,"percentile_exact":{"denominator":77,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647}],"result_updated_at":"2026-08-26","score":40.07,"score_exact":{"denominator":111251564424,"numerator":4457774361283}}],"release_date":"2026-06-01","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.7-plus"},"m-724c2a784624afa2":{"id":"m-724c2a784624afa2","in_window":true,"name":"Ling-3.0-flash-VL","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","model_id":"m-724c2a784624afa2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","model_id":"m-724c2a784624afa2","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","model_id":"m-724c2a784624afa2","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","model_id":"m-724c2a784624afa2","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":531,"numerator":21650},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3500},"family":"GPQA Diamond","row_ids":["c2e927593d58bea9e466eb1677b5bc62"],"score":59.32},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":200},"family":"HLE","row_ids":["4fb23f5b07fa697e50fa64d17b095add"],"score":22.22}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":59.32,"percentile_exact":{"denominator":59,"numerator":3500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond","peer_count":64,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076}],"score":40.77,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":64571609},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":925},"family":"tau3-Banking","row_ids":["a5800dd77226525d5b0005d3b9d5c59c"],"score":71.15},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4100},"family":"GDPval-AA","row_ids":["2d97f2a57324a678b4a67b9a04a099de"],"score":50.62},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["b740ee2b3c6f41e732abe95d92f63313"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":650},"family":"AutomationBench","row_ids":["1aac376daa38513de1520f91177fbd00"],"score":34.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3100},"family":"AA-Briefcase","row_ids":["3a94cd016eb68403ef51f517cb4b5722"],"score":40.26}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":71.15,"percentile_exact":{"denominator":13,"numerator":925},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v2.1","peer_count":82,"percentile":50.62,"percentile_exact":{"denominator":81,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v1.1","peer_count":78,"percentile":40.26,"percentile_exact":{"denominator":77,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95}],"score":41.91,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":700},"family":"GDP.pdf","row_ids":["8a1843eddbf9ae3836989a362277ba86"],"score":36.84}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084}],"score":36.84,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":800},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":800},"family":"MMMU-Pro","row_ids":["487c57fe9e6c668745daa6efed20c475"],"score":47.06}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543}],"score":47.06,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":59.32,"percentile_exact":{"denominator":59,"numerator":3500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond","peer_count":64,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":71.15,"percentile_exact":{"denominator":13,"numerator":925},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v2.1","peer_count":82,"percentile":50.62,"percentile_exact":{"denominator":81,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v1.1","peer_count":78,"percentile":40.26,"percentile_exact":{"denominator":77,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543}],"result_updated_at":null,"score":40.93,"score_exact":{"denominator":24722569872,"numerator":1011958813885}}],"release_date":"2026-09-10","release_status":"published","release_verified":true,"source_id":"inclusionai/ling-3.0-flash-vl"},"m-72b806dc67894cf3":{"id":"m-72b806dc67894cf3","in_window":true,"name":"Mistral Large 4 Preview","observed_records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6ce21a92becd0e73a78db51e210dc9c","model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1187.74,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1187.74},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"aa43eca35d1d3e248b0781cd2014eb99","model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.225,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.225},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcba3c26b001fc8948c02459b7c35c1c","model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7518279569892473,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7518279569892473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95c4e92e95f3808239675253ef04a041","model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5280083768459665,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5280083768459665},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"22073903c1a5d3f86c5d22ec711f3b8e","model_id":"m-72b806dc67894cf3","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","model_id":"m-72b806dc67894cf3","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"5d9f5492ddeb8760b1939254ffdac3ab","model_id":"m-72b806dc67894cf3","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.133204545454545,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.133204545454545},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","model_id":"m-72b806dc67894cf3","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"752cc0e021ce8ff217727c8faa57f582","model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"de50d6cfad22ec4871139b02d15844ca","model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.168,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.168},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"daa4b480c35fe0c512f1d71535e5d9e8","model_id":"m-72b806dc67894cf3","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.232323232323232,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232323232323232},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","model_id":"m-72b806dc67894cf3","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6edf7b36ea7ff4ee19308760052a1d50","model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.225,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.225},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":27846,"numerator":1152575},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":50},"family":"FrontierMath","row_ids":["8462c2a29fbbd735e478ce804db0a935"],"score":5.56},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["9bdac139a55bab7bdafc52c1059a96f7"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":400},"family":"HLE","row_ids":["3c62c893b3f2b3d94df10089539fb20a"],"score":57.14},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2500},"family":"CritPt","row_ids":["e52491cdc38740f220e3951458e0d11e"],"score":49.02}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","peer_count":64,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","peer_count":52,"percentile":49.02,"percentile_exact":{"denominator":51,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27}],"score":41.39,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":13167,"numerator":878050},"families":[{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":200},"family":"Terminal-Bench","row_ids":["038de3c11daf5dbf02e5d1c9dd11ccd7"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1350},"family":"AutomationBench","row_ids":["7fa3942a02994b01f2c4e2365b0583bc"],"score":71.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4800},"family":"AA-Briefcase","row_ids":["bc12c3c00877323848c26ea730d935cc"],"score":62.34}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"4.0","peer_count":76,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":78,"percentile":62.34,"percentile_exact":{"denominator":77,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31}],"score":66.69,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":2200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2200},"family":"FrontierCode","row_ids":["c316e6a1765e379f023423782e08477b"],"score":44.9}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"2.0","peer_count":50,"percentile":44.9,"percentile_exact":{"denominator":49,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313}],"score":44.9,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2154950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1350},"family":"GDP.pdf","row_ids":["a6443ea346efcda5f00f1ce182e0d982"],"score":71.05},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":50},"family":"SimpleQA Verified","row_ids":["39e767a0ed9e71374bab635fd9e2ce92"],"score":2.94},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2750},"family":"AA-Omniscience","row_ids":["6656d06e85a16beb97db67e4960f17c9","8175922707070fb09367c5a6424e1c3a"],"score":51.89}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018}],"score":41.96,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","peer_count":64,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","peer_count":52,"percentile":49.02,"percentile_exact":{"denominator":51,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"4.0","peer_count":76,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":78,"percentile":62.34,"percentile_exact":{"denominator":77,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"2.0","peer_count":50,"percentile":44.9,"percentile_exact":{"denominator":49,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018}],"result_updated_at":"2026-10-12","score":49.32,"score_exact":{"denominator":38864717892,"numerator":1916929064525}}],"release_date":"2026-10-06","release_status":"published","release_verified":true,"source_id":"mistral/mistral-large-4"},"m-754df8b5922bf5da":{"id":"m-754df8b5922bf5da","in_window":true,"name":"Solar Pro 4","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","model_id":"m-754df8b5922bf5da","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","model_id":"m-754df8b5922bf5da","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649}],"official_evidence":{"artifact_sha256":"4c3c4258c97035773c61772893086ae86059b2a773780452e3c61615afc1a64f","basis":"official_release_page","date":"2026-08-11","sha256":"4c3c4258c97035773c61772893086ae86059b2a773780452e3c61615afc1a64f","url":"https://www.upstage.ai/blog/en/solar-pro-4"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":109575125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4100},"family":"GPQA Diamond","row_ids":["8e38fd38f9044aff514003f73b248716"],"score":69.49},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["a2c09e681871381cb89f80ae44384709"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":650},"family":"HLE","row_ids":["66e35510159ab65c6e04c83322148b65"],"score":10.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1850},"family":"CritPt","row_ids":["7de1f456199243fb6a03dcf75e8fd06c"],"score":36.27}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":69.49,"percentile_exact":{"denominator":59,"numerator":4100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18}],"score":33.35,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":58748329},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":725},"family":"tau3-Banking","row_ids":["87111691a6a525e89691d7ee783ff6e3"],"score":55.77},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3400},"family":"GDPval-AA","row_ids":["2235f722890b82a4475fc455a58c66ab"],"score":41.98},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["e5e132be2e915d3fe97356bbd06869f9"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":550},"family":"AutomationBench","row_ids":["ba2dbb617a983db0793fc0f1b006c6eb"],"score":28.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3900},"family":"AA-Briefcase","row_ids":["1caa339c8531983b8c2298b6a2b6d27c"],"score":50.65}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":55.77,"percentile_exact":{"denominator":13,"numerator":725},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","peer_count":82,"percentile":41.98,"percentile_exact":{"denominator":81,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":28.95,"percentile_exact":{"denominator":19,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":78,"percentile":50.65,"percentile_exact":{"denominator":77,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32}],"score":38.13,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":2014,"numerator":86325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":625},"family":"GDP.pdf","row_ids":["75cb1248936655503fa568c6e5bc1e02"],"score":32.89},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2800},"family":"AA-Omniscience","row_ids":["fcae50a41df6fb08e7a7f301689104ab","9364a9aa01cea849d2eb414760bb16ec"],"score":52.83}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255}],"score":42.86,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":69.49,"percentile_exact":{"denominator":59,"numerator":4100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":55.77,"percentile_exact":{"denominator":13,"numerator":725},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","peer_count":82,"percentile":41.98,"percentile_exact":{"denominator":81,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":28.95,"percentile_exact":{"denominator":19,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":78,"percentile":50.65,"percentile_exact":{"denominator":77,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255}],"result_updated_at":null,"score":37.78,"score_exact":{"denominator":4586036711256,"numerator":173241935325145}}],"release_date":"2026-08-11","release_status":"published","release_verified":true,"source_id":"upstage/solar-pro4"},"m-77b5dbc317075e9d":{"id":"m-77b5dbc317075e9d","in_window":true,"name":"Gemini 3.6 Flash","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"official_evidence":{"artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","basis":"official_release_page","date":"2026-07-21","sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["1632686674422c3a2065b7adb4e89430"],"score":31.58}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":31.58,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":200},"families":[{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":200},"family":"AA-Briefcase","row_ids":["ec2017cdf9726d6330e3b72963cbf545"],"score":22.22}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808}],"score":22.22,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["2125ffdac9125e27a313037f91e5edb9"],"score":25.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0}],"score":25.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1325},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1325},"family":"SimpleQA Verified","row_ids":["8f6725165e10f2ce17308791e15f9d1c"],"score":77.94}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662}],"score":77.94,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":500},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":500},"family":"Humanity Sixth Sense","row_ids":["6d9db035c8e86692a9eb736f0a2c02ca"],"score":55.56}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","peer_count":28,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9}],"score":55.56,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","peer_count":28,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9}],"result_updated_at":"2026-08-27","score":39.59,"score_exact":{"denominator":2907,"numerator":115100}}],"release_date":"2026-07-21","release_status":"published","release_verified":true,"source_id":"google/gemini-3.6-flash"},"m-7990b74b056a0b9a":{"id":"m-7990b74b056a0b9a","in_window":true,"name":"Celeris-1","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-24","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":57550},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":750},"family":"GPQA Diamond","row_ids":["63aced451500a12a72f3e068bcd6ff1f"],"score":12.71},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1150},"family":"HLE","row_ids":["04be87f5b8e53109685297a0e0936bc5"],"score":18.25}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057}],"score":15.48,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":118503,"numerator":588163},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":0},"family":"tau3-Banking","row_ids":["09489dd17969a18613dce0aff6c534ba"],"score":0.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":400},"family":"GDPval-AA","row_ids":["0753db930b01dc3794424662db360ca1"],"score":4.94},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["732ce0d4c4b93e00640afbdf97e81cd3"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":75},"family":"AutomationBench","row_ids":["b4943be7200f5ac57de9f15aacbca48d"],"score":3.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":200},"family":"AA-Briefcase","row_ids":["dda521b5d7d0c5ce2d8db12c4ea78f4c"],"score":2.6}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","peer_count":82,"percentile":4.94,"percentile_exact":{"denominator":81,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","peer_count":78,"percentile":2.6,"percentile_exact":{"denominator":77,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94}],"score":4.96,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":125},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":125},"family":"GDP.pdf","row_ids":["d70ece05a7cd9206504e0831b954b698"],"score":6.58}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012}],"score":6.58,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","peer_count":82,"percentile":4.94,"percentile_exact":{"denominator":81,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","peer_count":78,"percentile":2.6,"percentile_exact":{"denominator":77,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012}],"result_updated_at":null,"score":9.18,"score_exact":{"denominator":13983354,"numerator":128393905}}],"release_date":"2026-07-24","release_status":"published","release_verified":true,"source_id":"published/aa:celeris-1"},"m-7b2a8fe101678ebd":{"id":"m-7b2a8fe101678ebd","in_window":true,"name":"HyperNova 60B 2605","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":6,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":1400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1400},"family":"GPQA Diamond","row_ids":["bbee1bae6aa788cf47294ddc38f9761f"],"score":23.73}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":23.73,"percentile_exact":{"denominator":59,"numerator":1400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333}],"score":23.73,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":648648,"numerator":5385785},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":125},"family":"tau3-Banking","row_ids":["24158d9869238dabbd834ff7fd267dc4"],"score":4.81},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":800},"family":"GDPval-AA","row_ids":["fc44351a5b9d8e210a763f56506c0719"],"score":9.88},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["2b6eac192b734b825e56a7c642b341b7"],"score":13.33},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":400},"family":"AA-Briefcase","row_ids":["606c6cc32d7931469129472a4d6fffcd"],"score":5.19}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","peer_count":82,"percentile":9.88,"percentile_exact":{"denominator":81,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","peer_count":78,"percentile":5.19,"percentile_exact":{"denominator":77,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41}],"score":8.3,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4,"numerator":275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":275},"family":"IFBench","row_ids":["ea00f82eb447a348674a4f2b40a4a4f4"],"score":68.75}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136}],"score":68.75,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":23.73,"percentile_exact":{"denominator":59,"numerator":1400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","peer_count":82,"percentile":9.88,"percentile_exact":{"denominator":81,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","peer_count":78,"percentile":5.19,"percentile_exact":{"denominator":77,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136}],"result_updated_at":null,"score":31.08,"score_exact":{"denominator":535783248,"numerator":16653656375}}],"release_date":"2026-05-26","release_status":"published","release_verified":true,"source_id":"published/aa:hypernova-60b"},"m-7ff6a9f995dcaa1c":{"id":"m-7ff6a9f995dcaa1c","in_window":true,"name":"Gemini 4 Argon","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench 4.0 (AA)","benchmark_version":"4.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"39d29b69788dd7332642aeadf008dc34","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench 4.0 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"6dfc6fa538421f2fee6d35c4","raw_value":57.1,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"57.1%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench 4.0 (AA):Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench:Terminal-Bench 4.0 (AA)","source_record_id":"carbon-image:Terminal-Bench 4.0 (AA):Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":57.1},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench Zapier","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"b60faaa3ac13688ab0e4d3540dcbbf95","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published developer protocol","profile_key":"native","protocol_description":"Official comparison image; AutomationBench Zapier; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"f658b6663efcee3dc42343bd","raw_value":51.3,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"51.3%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:AutomationBench Zapier:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:AutomationBench:AutomationBench Zapier","source_record_id":"carbon-image:AutomationBench Zapier:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":51.3},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science 0.1 (AA)","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"reasoning","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"58e1da3fabc235dff575e99c81331c7e","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench-Science 0.1 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"4f191cd5eb9caf57e48bc0e7","raw_value":52.9,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"52.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench-Science 0.1 (AA):Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench-Science:Terminal-Bench-Science 0.1 (AA)","source_record_id":"carbon-image:Terminal-Bench-Science 0.1 (AA):Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":52.9},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-4-argon","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c2167057895c8758046c1d7e9165f23f","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":53.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f361ec2ed229577db9d6a4b9"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f361ec2ed229577db9d6a4b9","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":53.2},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e3cc310052e96951792e7addae519033","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":57.58,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"57.58%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row8","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":57.58},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"84ccb8733065be35fe6ce8fbe381732b","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.528571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-argon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.528571428571429},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875}],"official_evidence":{"artifact_sha256":"555ea742fc3bcfcd343ab416bbcdf209448d15c9f06f0240eb19c9875fdccf00","basis":"official_release_page","date":"2026-10-13","sha256":"555ea742fc3bcfcd343ab416bbcdf209448d15c9f06f0240eb19c9875fdccf00","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-argon/"},"primary_profile_id":"native","profiles":[{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":3393175},"families":[{"benchmark_version":"2","exact":{"denominator":3,"numerator":200},"family":"FrontierMath","row_ids":["1004f40613fd7110bf20d6c47c5dcedb"],"score":66.67},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1700},"family":"Terminal-Bench-Science","row_ids":["50ddd51f0edebe68ad08c74703710c38"],"score":89.47},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5800},"family":"HLE","row_ids":["b185a4fb3502b3f94e9e93582e0d93a3"],"score":92.06},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1450},"family":"CritPt","row_ids":["3130f94996ac2afcc2f6ea8e78d8c87e"],"score":85.29}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","peer_count":39,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","peer_count":64,"percentile":92.06,"percentile_exact":{"denominator":63,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":83.37,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":16353414,"numerator":1123315619},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6500},"family":"GDPval-AA","row_ids":["feff6b1ffb0f6cb8ff2773defec53d29"],"score":80.25},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":278},"family":"Terminal-Bench","row_ids":["58499c853c2644c7eb9a4bc16fc208c6"],"score":92.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1875},"family":"AutomationBench","row_ids":["a2113bd24aca42a48eba4972dbb6e6d4"],"score":98.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5800},"family":"AA-Briefcase","row_ids":["782453050df6541e218363a18f9bb044"],"score":75.32},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"OSWorld","row_ids":["f61750e8ba74bd2c7928dd7c56124cb4"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1500},"family":"AA-AnalystAgent","row_ids":["4f1262f25c1f33cd755aebe2e08018dd"],"score":65.22}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","peer_count":82,"percentile":80.25,"percentile_exact":{"denominator":81,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","peer_count":76,"percentile":92.67,"percentile_exact":{"denominator":3,"numerator":278},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":78,"percentile":75.32,"percentile_exact":{"denominator":77,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","peer_count":24,"percentile":65.22,"percentile_exact":{"denominator":23,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875}],"score":68.69,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":1911,"numerator":161950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3600},"family":"FrontierCode","row_ids":["aabef6305c7bf9e588cc8a4221b8394d"],"score":73.47},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1050},"family":"Vals Code Migration","row_ids":["2905dc672ed8db458e1dc8c5a09c04a7"],"score":80.77},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":100},"family":"MirrorCode","row_ids":["d2afdbf96101f2fb4499bd4ee1fc2881"],"score":100.0}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","peer_count":50,"percentile":73.47,"percentile_exact":{"denominator":49,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","peer_count":27,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889}],"score":84.75,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":2014,"numerator":175925},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1575},"family":"GDP.pdf","row_ids":["da4cc7b10c537dba69aca5448e60f8f7"],"score":82.89},{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":100},"family":"SimpleQA Verified","row_ids":["56d131d2a5cbdffe166bdda67e598654"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"MultiChallenge","row_ids":["2db7888810c9c42969f9152c713f9d6e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4850},"family":"AA-Omniscience","row_ids":["0588c4d7908c2d87deb09a5a03cbe1c3","2f9d0f30b6d18925324be8097217f20f"],"score":91.51}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383}],"score":87.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":2,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["18175febe04f7246f991614fc6ca93f9"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"Chartography","row_ids":["13ad1c51a1b24d02661158660b44e5e0"],"score":75.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","peer_count":28,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6}],"score":87.5,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","peer_count":39,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","peer_count":64,"percentile":92.06,"percentile_exact":{"denominator":63,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","peer_count":82,"percentile":80.25,"percentile_exact":{"denominator":81,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","peer_count":76,"percentile":92.67,"percentile_exact":{"denominator":3,"numerator":278},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":78,"percentile":75.32,"percentile_exact":{"denominator":77,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","peer_count":24,"percentile":65.22,"percentile_exact":{"denominator":23,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","peer_count":50,"percentile":73.47,"percentile_exact":{"denominator":49,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","peer_count":27,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","peer_count":28,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6}],"result_updated_at":"2026-10-28","score":81.19,"score_exact":{"denominator":2681665534548,"numerator":217712535238087}}],"release_date":"2026-10-13","release_status":"published","release_verified":true,"source_id":"google/gemini-4-argon"},"m-825009614c016c1d":{"id":"m-825009614c016c1d","in_window":true,"name":"Qwen3.8 Max (0902)","observed_records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"ec578f1a940281fdcb27687071dac7ac","model_id":"m-825009614c016c1d","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":34.0,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:Qwen3.8 Max (0902)"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:Qwen3.8 Max (0902)","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","model_id":"m-825009614c016c1d","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","model_id":"m-825009614c016c1d","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","model_id":"m-825009614c016c1d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","model_id":"m-825009614c016c1d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","model_id":"m-825009614c016c1d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","model_id":"m-825009614c016c1d","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","model_id":"m-825009614c016c1d","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","model_id":"m-825009614c016c1d","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"901770e3927f8efbffa8b7b6dc69571d","model_id":"m-825009614c016c1d","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.388888888888889,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-max:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.388888888888889},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","model_id":"m-825009614c016c1d","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","model_id":"m-825009614c016c1d","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","model_id":"m-825009614c016c1d","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":821457,"numerator":59247200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5350},"family":"GPQA Diamond","row_ids":["65e0dc8726d7a3e7afde224768d43963"],"score":90.68},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":950},"family":"Terminal-Bench-Science","row_ids":["4ea60fe9901ae1fc5d3c1daf559ab6e1"],"score":73.08},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4400},"family":"HLE","row_ids":["2ab50d9a5a2d86ccb853771d429875d3"],"score":69.84},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["b1d4b1cfdd4d38290545201e6c39db63"],"score":54.9}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","included":true,"model_id":"m-825009614c016c1d","original_version_label":"0.1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond","peer_count":64,"percentile":69.84,"percentile_exact":{"denominator":63,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":72.12,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":2267673408,"numerator":186901243525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["ae1b88ab2d0880e90cfd75002e9f4d77"],"score":100.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1225},"family":"tau3-Banking","row_ids":["1cd94c3f93e80676324cc2282ca34aa4"],"score":94.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7100},"family":"GDPval-AA","row_ids":["8fccbdc41b807b2ebbe41cff024b0690"],"score":87.65},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1675},"family":"Terminal-Bench","row_ids":["0d2e487f06bf23cbe7ee7f90f6003d16"],"score":69.79},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1275},"family":"AutomationBench","row_ids":["308f05669cd695ca55d1d348f97d3b83"],"score":67.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6900},"family":"AA-Briefcase","row_ids":["ecc6d7d48519d5a999fc5f71c76c0b6d"],"score":89.61},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":850},"family":"EnterpriseOps-Gym","row_ids":["aaa4039880a2e3c56a0fee65a0cf496e"],"score":94.44},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1300},"family":"AA-AnalystAgent","row_ids":["cee5a16789123af6ef334475ebff73a7"],"score":56.52}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","included":true,"model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":94.23,"percentile_exact":{"denominator":13,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2.1","peer_count":82,"percentile":87.65,"percentile_exact":{"denominator":81,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","included":true,"model_id":"m-825009614c016c1d","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","included":true,"model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":67.11,"percentile_exact":{"denominator":19,"numerator":1275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":78,"percentile":89.61,"percentile_exact":{"denominator":77,"numerator":6900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","peer_count":24,"percentile":56.52,"percentile_exact":{"denominator":23,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45}],"score":82.42,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":40975},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3100},"family":"FrontierCode","row_ids":["6cbe0b4839fdc145889ebffe2def2a50"],"score":63.27},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":850},"family":"Vals Code Migration","row_ids":["62c5b15d1c0217ab796423502244a099"],"score":65.38}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","included":true,"model_id":"m-825009614c016c1d","original_version_label":"2.0","peer_count":50,"percentile":63.27,"percentile_exact":{"denominator":49,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","included":true,"model_id":"m-825009614c016c1d","original_version_label":"1.1","peer_count":27,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19}],"score":64.32,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2127575},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3225},"family":"GDP.pdf","row_ids":["5928f02dfd05733dd4d23f64991980e9"],"score":84.87},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":700},"family":"SimpleQA Verified","row_ids":["dd26bef806a0bba267bbd1b02770b43f"],"score":41.18},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3200},"family":"AA-Omniscience","row_ids":["c42d799db980160b63c3bab31d80fa02","602505b7f91e3456e6da6058f7b120bf"],"score":60.38}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","included":true,"model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088}],"score":62.14,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1300},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1300},"family":"MMMU-Pro","row_ids":["502735168b948696e5ae04c6357e58b6"],"score":76.47}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","peer_count":35,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884}],"score":76.47,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","included":true,"model_id":"m-825009614c016c1d","original_version_label":"0.1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond","peer_count":64,"percentile":69.84,"percentile_exact":{"denominator":63,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","included":true,"model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":94.23,"percentile_exact":{"denominator":13,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2.1","peer_count":82,"percentile":87.65,"percentile_exact":{"denominator":81,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","included":true,"model_id":"m-825009614c016c1d","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","included":true,"model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":67.11,"percentile_exact":{"denominator":19,"numerator":1275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":78,"percentile":89.61,"percentile_exact":{"denominator":77,"numerator":6900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","peer_count":24,"percentile":56.52,"percentile_exact":{"denominator":23,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","included":true,"model_id":"m-825009614c016c1d","original_version_label":"2.0","peer_count":50,"percentile":63.27,"percentile_exact":{"denominator":49,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","included":true,"model_id":"m-825009614c016c1d","original_version_label":"1.1","peer_count":27,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","included":true,"model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","peer_count":35,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884}],"result_updated_at":"2026-09-02","score":71.58,"score_exact":{"denominator":3375323019484416,"numerator":241593097629493205}}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.8-max-0902"},"m-8555d29b166d94e4":{"id":"m-8555d29b166d94e4","in_window":true,"name":"Nanbeige4.3-33B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","model_id":"m-8555d29b166d94e4","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","model_id":"m-8555d29b166d94e4","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","model_id":"m-8555d29b166d94e4","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155}],"official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-10-25","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":2457,"numerator":80425},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":575},"family":"Terminal-Bench-Science","row_ids":["63428ab4d6f5aa0cb6c23626f378d1a8"],"score":44.23},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1300},"family":"HLE","row_ids":["5c428eec7c6c7eddbfd00d536e4c24e0"],"score":20.63},{"benchmark_version":"2 Hard","exact":{"denominator":3,"numerator":100},"family":"CritPt","row_ids":["2147a64a171deda69e1dcacdea28ad28"],"score":33.33}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"0.1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","peer_count":64,"percentile":20.63,"percentile_exact":{"denominator":63,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","peer_count":52,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155}],"score":32.73,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":4389,"numerator":275636},"families":[{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":72},"family":"Terminal-Bench","row_ids":["37a1dedbc40bda682abb927c484863ce"],"score":72.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1200},"family":"AutomationBench","row_ids":["87eb4175855e40447de186dd2f3411a6"],"score":63.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4100},"family":"AA-Briefcase","row_ids":["24d71c9e7d3f6a3d2ddf4b46e341dafa"],"score":53.25}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"4.0","peer_count":76,"percentile":72.0,"percentile_exact":{"denominator":1,"numerator":72},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":78,"percentile":53.25,"percentile_exact":{"denominator":77,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84}],"score":62.8,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1800},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1800},"family":"FrontierCode","row_ids":["9302ab59d95f9c2a82c7d7c38865f23b"],"score":36.73}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"2.0","peer_count":50,"percentile":36.73,"percentile_exact":{"denominator":49,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202}],"score":36.73,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":195525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["632aef0c9b2c1ee60a8fee169a7ead99"],"score":48.03},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2600},"family":"AA-Omniscience","row_ids":["464c122b5f2d778a04114dc266e1a6b8","d5ec18e90fffd664c6089d00e270f823"],"score":49.06}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698}],"score":48.54,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"0.1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","peer_count":64,"percentile":20.63,"percentile_exact":{"denominator":63,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","peer_count":52,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"4.0","peer_count":76,"percentile":72.0,"percentile_exact":{"denominator":1,"numerator":72},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":78,"percentile":53.25,"percentile_exact":{"denominator":77,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"2.0","peer_count":50,"percentile":36.73,"percentile_exact":{"denominator":49,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698}],"result_updated_at":null,"score":45.49,"score_exact":{"denominator":1714619907,"numerator":77994001630}}],"release_date":"2026-10-25","release_status":"published","release_verified":true,"source_id":"published/aa:nanbeige4-3-33b"},"m-86c0c8bf10aa4a0d":{"id":"m-86c0c8bf10aa4a0d","in_window":true,"name":"Claude Fable 5","observed_records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","id":"c935b5fe18707c971521047b772f70d1","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":17.4,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":17.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","id":"be49aacf10e50a43b90ee123c1c07cdd","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"ac0f80ad7501f371bea4ff3e","raw_value":44.5,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row1:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench:Terminal-Bench 4.0","source_record_id":"gpt6-astra-official:table2:row1:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":44.5},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"fe5b288d983212a01455a7bed99bcbe3","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench Science 0.1; preserve page footnotes and task-specific settings","protocol_id":"6e315c8a4731951529a013b5","raw_value":21.4,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row1:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench-Science:Terminal-Bench Science 0.1","source_record_id":"gpt6-astra-official:table3:row1:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":21.4},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"ca71de5904017749aaf06e4425e6f606","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.7,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"663672da9a7ca9303639232126d54075","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":87.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column7"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column7","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":87.0},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"15d1d7ba1938d406176f857ad1e590e4","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column7"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column7","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":87.8},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","id":"49505e077334acf60faa1c56e592c1f9","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":17.4,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column7"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column7","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":17.4},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"0a11f2713c3020fc91ed146a0cde7d3d","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.6%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column7"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column7","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"99378a4787b7b2fe45ecccbd3c678fc6","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":70.0,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"70.0","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":70.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"e6769544adbda039071845c194ca792d","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":29.1,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"29.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":29.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d42da14e4f5c6f46bc80f80cc52a8e79","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Fable5. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.6,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Fable 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"985255818b7c5bb489a1bbeb1bc94859","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":70.0,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"70.0","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":70.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"e5e10458ba007e3ac9010def1e16a95d","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":29.1,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"29.1","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":29.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Fable 5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"93d9797cec6ccd384ba684d52f6088bc","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":63.5,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"63.5","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":63.5},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Fable 5","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"6fcca244b7e23fe3074831e7ca3cf1be","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"f8d14a3afa2d20c0373188e9c415dc6c","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1; explicit developer footnote","profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":70.0,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"70.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":70.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"127d7af68dabb544e18741601786e437","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":29.1,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"29.1","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":29.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"f6eb2e4c060926f84692b3351b0182f9","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.2,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.2 / 86.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e90baba8f5537951896a42c0cc814488","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"92.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":92.6},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"fdab9cc10508c8861689ba5d61adcb18","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":84.7,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"84.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":84.7},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"34ae78b3ccf7f2237d26759361270f8d","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic scientific research Terminal-Bench-Science 0.1 [1]; preserve page footnotes and task-specific settings","protocol_id":"88023c6718be09c275d82cfc","raw_value":24.7,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"24.7%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row2:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1 [1]","source_record_id":"anthropic-fable51:table0:row2:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":24.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"3150cf0f12ca60dc3d5a4b9a023f31b8","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":42.0,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"42.0%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row4:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"anthropic-fable51:table0:row4:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":42.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71}],"official_evidence":{"artifact_sha256":"c8267145b47f034cb45aedbe8074a7b244fee4ab2050099a2f6e2e7f9d6d9ed5","basis":"official_release_page","date":"2026-06-09","sha256":"c8267145b47f034cb45aedbe8074a7b244fee4ab2050099a2f6e2e7f9d6d9ed5","url":"https://www.anthropic.com/news/claude-fable-5-mythos-5"},"primary_profile_id":"native","profiles":[{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":399,"numerator":23825},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":600},"family":"FrontierMath","row_ids":["51cb1dea03ece9a899f843aedba7597e"],"score":85.71},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":25},"family":"GPQA Diamond","row_ids":["b6943c0f15a31bf97046c2e4af0b90f3"],"score":25.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1300},"family":"Terminal-Bench-Science","row_ids":["f5cc9d3986b503f55588d97bf5fde40a"],"score":68.42}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","peer_count":8,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","peer_count":39,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71}],"score":59.71,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":144,"numerator":8225},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["b82b11faf1fed2434465112eb548316c"],"score":100.0},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":875},"family":"Terminal-Bench","row_ids":["5ecd3796946e727694a868a88027022a"],"score":72.92},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["8ad20dc033b0a09d811fcebfca51c2ab"],"score":0.0},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":500},"family":"MCP Atlas","row_ids":["37801ada9ba8d94f8b7e8f5a6667ee11"],"score":55.56}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","peer_count":49,"percentile":72.92,"percentile_exact":{"denominator":12,"numerator":875},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3}],"score":57.12,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"DeepSWE","row_ids":["0772cd4ebc7a33c01cf749b6a69bd7c0"],"score":50.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","peer_count":9,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9}],"score":50.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":2,"numerator":95},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"IFBench","row_ids":["3f5e26906c244a4d4f9e57725da3d9df"],"score":20.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"GDP.pdf","row_ids":["134a991b936b3bccddf2a9a60c450e69"],"score":75.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8}],"score":47.5,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":3,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["cf924b68bd130c394ca909ce8970cb90"],"score":66.67},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["a6ccc017ec5976d43545f1578c4ba3f8"],"score":50.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","peer_count":28,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2}],"score":58.33,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","peer_count":8,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","peer_count":39,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","peer_count":49,"percentile":72.92,"percentile_exact":{"denominator":12,"numerator":875},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","peer_count":9,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","peer_count":28,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2}],"result_updated_at":"2026-09-29","score":54.54,"score_exact":{"denominator":76608,"numerator":4178261}}],"release_date":"2026-06-09","release_status":"published","release_verified":true,"source_id":"anthropic/claude-fable-5"},"m-8877a9675c79fcad":{"id":"m-8877a9675c79fcad","in_window":true,"name":"Command A+","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d233621c7914e527a1ce47017fbfa4dd","model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.739455782312925,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"command-a-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.739455782312925},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","model_id":"m-8877a9675c79fcad","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","model_id":"m-8877a9675c79fcad","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","model_id":"m-8877a9675c79fcad","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"36e54b0974d1c45e6ed2c1cfd178e9b4","model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.527272727272727,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.527272727272727},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-20","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":67424575},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2100},"family":"GPQA Diamond","row_ids":["13075042d3067db5be44e3af38eac3de"],"score":35.59},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["705b0d6767e14eda4ee6da31459437c6"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":850},"family":"HLE","row_ids":["48d08d1a4f13b1f74f0a43f0e70c78cb"],"score":13.49},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":800},"family":"CritPt","row_ids":["bfeb15a230a3fe93a4b8668ae5c71e33"],"score":15.69}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":35.59,"percentile_exact":{"denominator":59,"numerator":2100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","peer_count":52,"percentile":15.69,"percentile_exact":{"denominator":51,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025}],"score":20.52,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2567565,"numerator":38748266},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":100},"family":"tau3-Banking","row_ids":["d4ac6d97f1d2caee1aac75fe86cb1d41"],"score":7.69},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":400},"family":"GDPval-AA","row_ids":["a7f8996adb7838234a9dd47a6d848a6c"],"score":14.81},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["ed0927341bbf29454c5f15ec4d2315cd"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":250},"family":"AutomationBench","row_ids":["c9594ae7e63800079a6fe5e87536af10"],"score":13.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":600},"family":"AA-Briefcase","row_ids":["d623fd47b28047c5307a67919c5c6fae"],"score":7.79}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","peer_count":82,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":13.16,"percentile_exact":{"denominator":19,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":78,"percentile":7.79,"percentile_exact":{"denominator":77,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2}],"score":15.09,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["c8987663cd611e1b09bc6211e110c938"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":81325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":25},"family":"IFBench","row_ids":["3f2a6e6ace968612fc72d2719d04f5ee"],"score":6.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":100},"family":"GDP.pdf","row_ids":["6612c73a65ff838439bc02c707904f4b"],"score":5.26},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2600},"family":"AA-Omniscience","row_ids":["0ca7e9728f633cc599fad0ed12a6097b","862933c1b9c8be1d0cbb4bb130b222e9"],"score":49.06}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","peer_count":17,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625}],"score":20.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":50},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":50},"family":"MMMU-Pro","row_ids":["b79c49748f47d291a424306d40dc6469"],"score":2.94}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532}],"score":2.94,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":35.59,"percentile_exact":{"denominator":59,"numerator":2100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","peer_count":52,"percentile":15.69,"percentile_exact":{"denominator":51,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","peer_count":82,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":13.16,"percentile_exact":{"denominator":19,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":78,"percentile":7.79,"percentile_exact":{"denominator":77,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","peer_count":17,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532}],"result_updated_at":null,"score":13.85,"score_exact":{"denominator":15286789037520,"numerator":211678630100407}}],"release_date":"2026-05-20","release_status":"published","release_verified":true,"source_id":"cohere/command-a-plus-05-2026"},"m-8d99798a7e24690c":{"id":"m-8d99798a7e24690c","in_window":true,"name":"Gemini 3.5 Flash","observed_records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","model_id":"m-8d99798a7e24690c","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:14.844438+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-669a69be0c5e6f05","id":"cb083b6eb13cb586270ac912934c99eb","model_id":"m-8d99798a7e24690c","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page google35-flash-card; row MMMU-Pro Multimodal understanding and reasoning; preserve page footnotes and task-specific settings","protocol_id":"a1499705a4d30b9b15b31b15","raw_value":83.6,"research_artifact_sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","response_sha256":"3b5bb38ac75ac4a7b458e5d3ae04109513f1049cdadca164149c9eae57ed0b6c","result_date":"2026-05-19","source_cell":"83.6%","source_id":"s-5f0593ef47a5116017ed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"google35-flash-card:table0:row9:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:MMMU-Pro:MMMU-Pro Multimodal understanding and reasoning","source_record_id":"google35-flash-card:table0:row9:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-5-flash/","unit":"%","value":83.6},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"dcee7e42fae3af18a3a914edeb2ea99e","model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2773207990599295,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2fc75bb992ddc7d12688edb3"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"2fc75bb992ddc7d12688edb3","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2773207990599295},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","model_id":"m-8d99798a7e24690c","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"official_evidence":{"artifact_sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","basis":"official_release_page","date":"2026-05-19","sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","url":"https://deepmind.google/models/model-cards/gemini-3-5-flash/"},"primary_profile_id":"native","profiles":[{"benchmark_count":8,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":850},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["72ae1d1de2eb58dbc8e135c9056f08ba"],"score":44.74}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"score":44.74,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":96,"numerator":2275},"families":[{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":275},"family":"Terminal-Bench","row_ids":["349a8da908aa998fbc2c641cf58cfc1b"],"score":11.46},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["d209f552e4c2026bb776833fb9d113ca"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":50},"family":"AA-Briefcase","row_ids":["a4acfc86f781d098a642bc8bcfd57f42"],"score":16.67},{"benchmark_version":"1000 tasks","exact":{"denominator":3,"numerator":200},"family":"MCP Atlas","row_ids":["44434b3879169ad5c54a4316dc8731a9"],"score":66.67}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6}],"score":23.7,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["7b14fb9c23111ece0c4ecf6bf486054c"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1325},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1325},"family":"SimpleQA Verified","row_ids":["f90868b4246f3217c1fa61fa9b95b789"],"score":77.94}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662}],"score":77.94,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":27,"numerator":1600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1600},"family":"Humanity Sixth Sense","row_ids":["1fedb85017624b14b5d56927cae80633"],"score":59.26}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","peer_count":28,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6}],"score":59.26,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","peer_count":28,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6}],"result_updated_at":"2026-08-27","score":38.62,"score_exact":{"denominator":1116288,"numerator":43114225}}],"release_date":"2026-05-19","release_status":"published","release_verified":true,"source_id":"google/gemini-3.5-flash"},"m-8e76831a5814eec9":{"id":"m-8e76831a5814eec9","in_window":true,"name":"MiniMax-M3.1-Flash","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3aab4755f21529c39094e459048b22d6","model_id":"m-8e76831a5814eec9","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":25.25,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"25.25%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":25.25},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","model_id":"m-8e76831a5814eec9","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","model_id":"m-8e76831a5814eec9","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","model_id":"m-8e76831a5814eec9","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-19","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1071,"numerator":66400},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":50},"family":"Terminal-Bench-Science","row_ids":["e374e482dae499c526f6a7da9baef1aa"],"score":50.0},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1600},"family":"HLE","row_ids":["a1e07609863e9a13d3798af4a816cb48"],"score":76.19},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3050},"family":"CritPt","row_ids":["321a79f217d130c0ceb39102c3c1fab9"],"score":59.8}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"0.1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","peer_count":64,"percentile":76.19,"percentile_exact":{"denominator":21,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325}],"score":62.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":474012,"numerator":36496535},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5500},"family":"GDPval-AA","row_ids":["cfa7d9348c146c713f91a2534eae2b29"],"score":67.9},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":70},"family":"Terminal-Bench","row_ids":["3a8fd5af00baca1a19eb582e1a8c4ab5"],"score":70.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1825},"family":"AutomationBench","row_ids":["b04984f7d9488e18c36d59811ea3856b"],"score":96.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5700},"family":"AA-Briefcase","row_ids":["da58b11d089faf8c31c386386035abd7"],"score":74.03}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","peer_count":82,"percentile":67.9,"percentile_exact":{"denominator":81,"numerator":5500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":96.05,"percentile_exact":{"denominator":19,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":78,"percentile":74.03,"percentile_exact":{"denominator":77,"numerator":5700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98}],"score":76.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":2700},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2700},"family":"FrontierCode","row_ids":["1935649843d5e809cef231b07734be32"],"score":55.1}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"2.0","peer_count":50,"percentile":55.1,"percentile_exact":{"denominator":49,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182}],"score":55.1,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":237475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2975},"family":"GDP.pdf","row_ids":["96a35bebfd1cf232725317d8f36f4c53"],"score":78.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["2ab90fc472453676cb1037409906c906","cd0ef2b66ec6818e014beedc4ad88c9f"],"score":39.62}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483}],"score":58.96,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":900},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":900},"family":"MMMU-Pro","row_ids":["fc2b137ca09cf3cd0123e67faaeeed34"],"score":52.94}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786}],"score":52.94,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"0.1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","peer_count":64,"percentile":76.19,"percentile_exact":{"denominator":21,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","peer_count":82,"percentile":67.9,"percentile_exact":{"denominator":81,"numerator":5500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":96.05,"percentile_exact":{"denominator":19,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":78,"percentile":74.03,"percentile_exact":{"denominator":77,"numerator":5700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"2.0","peer_count":50,"percentile":55.1,"percentile_exact":{"denominator":49,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786}],"result_updated_at":null,"score":62.85,"score_exact":{"denominator":703433808,"numerator":44213635945}}],"release_date":"2026-10-19","release_status":"published","release_verified":true,"source_id":"minimax/MiniMax-M3.1-Flash"},"m-928c410ba1082113":{"id":"m-928c410ba1082113","in_window":true,"name":"K2 Horizon 0.9B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","model_id":"m-928c410ba1082113","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","model_id":"m-928c410ba1082113","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","model_id":"m-928c410ba1082113","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":100},"family":"GPQA Diamond","row_ids":["990a7e694606915f66b15d0a96c06abb"],"score":1.69}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","included":true,"model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":1.69,"percentile_exact":{"denominator":59,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293}],"score":1.69,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":39,"numerator":119},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":25},"family":"tau3-Banking","row_ids":["57be0e2f641053db7912c66d73e8ba7b"],"score":1.92},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"GDPval-AA","row_ids":["710c23bbfe14fa2ea7bd1f3a24100a69"],"score":0.0},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["7ed61dbea4d39e1cec33826bfcfe4762"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["e01a487a4aa0d6109e212abc8d8c200e"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"AA-Briefcase","row_ids":["3f053e6933f1e5ef77b0bf87fdfb5476"],"score":0.0}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","included":true,"model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":1.92,"percentile_exact":{"denominator":13,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v2.1","peer_count":82,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","included":true,"model_id":"m-928c410ba1082113","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","included":true,"model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v1.1","peer_count":78,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0}],"score":3.05,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":125},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":125},"family":"GDP.pdf","row_ids":["b9fa019c4e1362fc93bc4088a9dc8dbc"],"score":3.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","included":true,"model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008}],"score":3.29,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","included":true,"model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":1.69,"percentile_exact":{"denominator":59,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","included":true,"model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":1.92,"percentile_exact":{"denominator":13,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v2.1","peer_count":82,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","included":true,"model_id":"m-928c410ba1082113","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","included":true,"model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v1.1","peer_count":78,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","included":true,"model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008}],"result_updated_at":null,"score":2.63,"score_exact":{"denominator":612066,"numerator":1612745}}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"source_id":"published/aa:k2-horizon-0-9b"},"m-9453d55ff407aaee":{"id":"m-9453d55ff407aaee","in_window":true,"name":"Granite 4.2 30B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","model_id":"m-9453d55ff407aaee","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","model_id":"m-9453d55ff407aaee","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":41628025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":900},"family":"GPQA Diamond","row_ids":["651c19da954966c3a71036c8d3db174d"],"score":15.25},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["47e73ec519c6695efcb0b041ef5795ce"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":400},"family":"HLE","row_ids":["db2ce17572462c1eb76d15ea72c97e2b"],"score":6.35},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["d5d41ba5c0f73bb7b22fdca4916cb1b7"],"score":11.76}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":15.25,"percentile_exact":{"denominator":59,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","peer_count":64,"percentile":6.35,"percentile_exact":{"denominator":63,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":12.67,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":118503,"numerator":1813883},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":25},"family":"tau3-Banking","row_ids":["767893d84c1938b3d363a56f8204eb3b"],"score":25.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1400},"family":"GDPval-AA","row_ids":["a655895858c8427a30e33e8793a2117d"],"score":17.28},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["027860a90cc2579f5f5af8c36e50c87e"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":200},"family":"AutomationBench","row_ids":["354c0d77cbcd6264333d316fdf9f676b"],"score":10.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":800},"family":"AA-Briefcase","row_ids":["afdcb1efb0aca22bea9c16a3d016167d"],"score":10.39}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","peer_count":82,"percentile":17.28,"percentile_exact":{"denominator":81,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":10.53,"percentile_exact":{"denominator":19,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":78,"percentile":10.39,"percentile_exact":{"denominator":77,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75}],"score":15.31,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":107275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":375},"family":"GDP.pdf","row_ids":["6e95a209bd026922387753b507be18e1"],"score":9.87},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["d19df36b3cd3a0e840c1256cd1c744fe","82eab72497730b1d30475a9ca4c7819b"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948}],"score":26.63,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":15.25,"percentile_exact":{"denominator":59,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","peer_count":64,"percentile":6.35,"percentile_exact":{"denominator":63,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","peer_count":82,"percentile":17.28,"percentile_exact":{"denominator":81,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":10.53,"percentile_exact":{"denominator":19,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":78,"percentile":10.39,"percentile_exact":{"denominator":77,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948}],"result_updated_at":null,"score":17.6,"score_exact":{"denominator":4586036711256,"numerator":80716639816145}}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"source_id":"published/aa:granite-4-2-30b"},"m-95ce1c534519e66a":{"id":"m-95ce1c534519e66a","in_window":true,"name":"Ling-3.0-flash-Fin","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","model_id":"m-95ce1c534519e66a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":63,"numerator":1000},"families":[{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1000},"family":"HLE","row_ids":["11ffdf294b4e40190b29f471a44eb232"],"score":15.87}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","peer_count":64,"percentile":15.87,"percentile_exact":{"denominator":63,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":15.87,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":248026779,"numerator":8390522060},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"APEX-Agents","row_ids":["ca1ee026c3875e1f8a6f67a549e74fe8"],"score":25.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1000},"family":"tau3-Banking","row_ids":["c993b26e42ea6045232f1d9e621ef2fa"],"score":76.92},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3500},"family":"GDPval-AA","row_ids":["e73dd8b4c1ab25719febf181d0bca142"],"score":43.21},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["d43f0f17f32dc60cbed0bdb79fe20eb1"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":525},"family":"AutomationBench","row_ids":["7fbcd0afe5186232a5dba432e082b6ef"],"score":27.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2900},"family":"AA-Briefcase","row_ids":["11eea946a9e871701b9adaec2c3d14b2"],"score":37.66},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":300},"family":"AA-AnalystAgent","row_ids":["c283db79a5fa78b491d2c36ac505fd73"],"score":13.04}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","peer_count":82,"percentile":43.21,"percentile_exact":{"denominator":81,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":27.63,"percentile_exact":{"denominator":19,"numerator":525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","peer_count":78,"percentile":37.66,"percentile_exact":{"denominator":77,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","peer_count":24,"percentile":13.04,"percentile_exact":{"denominator":23,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625}],"score":33.83,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1075},"family":"GDP.pdf","row_ids":["0e26cf9d2beafb47ed1ed89556040c14"],"score":28.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":28.29,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","peer_count":64,"percentile":15.87,"percentile_exact":{"denominator":63,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","peer_count":82,"percentile":43.21,"percentile_exact":{"denominator":81,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":27.63,"percentile_exact":{"denominator":19,"numerator":525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","peer_count":78,"percentile":37.66,"percentile_exact":{"denominator":77,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","peer_count":24,"percentile":13.04,"percentile_exact":{"denominator":23,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"result_updated_at":null,"score":25.83,"score_exact":{"denominator":1736187453,"numerator":44851731725}}],"release_date":"2026-09-11","release_status":"published","release_verified":true,"source_id":"inclusionai/ling-3.0-flash-fin"},"m-98ee7005cf45b532":{"id":"m-98ee7005cf45b532","in_window":true,"name":"Claude Fable 5.1","observed_records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"agents","direction":"higher_better","id":"2372d74bdc501d7556cf0aa74944c192","model_id":"m-98ee7005cf45b532","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":31.4,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":31.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"agents","direction":"higher_better","id":"ceb85c397387a4847b40dd8b3d454d6e","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"ac0f80ad7501f371bea4ff3e","raw_value":55.8,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench:Terminal-Bench 4.0","source_record_id":"gpt6-astra-official:table2:row1:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":55.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","model_id":"m-98ee7005cf45b532","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"e9ea1a0c9a1ad3ceb806b63de3f0f7a8","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row Terminal-Bench Science 0.1; preserve page footnotes and task-specific settings","protocol_id":"6e315c8a4731951529a013b5","raw_value":52.6,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:Terminal-Bench-Science:Terminal-Bench Science 0.1","source_record_id":"gpt6-astra-official:table3:row1:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":52.6},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","model_id":"m-98ee7005cf45b532","original_version_label":"v2","profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"7d3842228f84e6a5e57e7db96b8839e3","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5.1. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench 4.0 (AA)","benchmark_version":"4.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"fc1ff1f2174461295bf4f514896624b4","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench 4.0 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"6dfc6fa538421f2fee6d35c4","raw_value":52.0,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"52.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench 4.0 (AA):Claude Fable 5.1"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench:Terminal-Bench 4.0 (AA)","source_record_id":"carbon-image:Terminal-Bench 4.0 (AA):Claude Fable 5.1","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":52.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"84a17f8a2fe109b449e2d00efd604380","model_id":"m-98ee7005cf45b532","original_version_label":"2.1","profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":76.8,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"76.8%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Claude Fable 5.1"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Claude Fable 5.1","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":76.8},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench Zapier","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"92973184577a9ca58d3865a6bc131f18","model_id":"m-98ee7005cf45b532","original_version_label":"current published developer protocol","profile_key":"native","protocol_description":"Official comparison image; AutomationBench Zapier; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"f658b6663efcee3dc42343bd","raw_value":31.4,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"31.4%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:AutomationBench Zapier:Claude Fable 5.1"},"source_name":"blog.google","source_protocol_key":"official:blog.google:AutomationBench:AutomationBench Zapier","source_record_id":"carbon-image:AutomationBench Zapier:Claude Fable 5.1","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":31.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"cee957e5db92decea0ed1ccab6ac32b5","model_id":"m-98ee7005cf45b532","original_version_label":"current published dataset","profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":46.2,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"46.2%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Claude Fable 5.1"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Claude Fable 5.1","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":46.2},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science 0.1 (AA)","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","dimension":"reasoning","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"11ef98afa95070c53ee2bc064c7c7936","model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench-Science 0.1 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"4f191cd5eb9caf57e48bc0e7","raw_value":43.3,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"43.3%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench-Science 0.1 (AA):Claude Fable 5.1"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench-Science:Terminal-Bench-Science 0.1 (AA)","source_record_id":"carbon-image:Terminal-Bench-Science 0.1 (AA):Claude Fable 5.1","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":43.3},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"c754242c94abd65cff43bbb52150d354","model_id":"m-98ee7005cf45b532","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1735.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1735","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1735.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"0d7edb34eeb6882f692808525c3c2a73","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Agentic scientific research Terminal-Bench-Science 0.1³; preserve page footnotes and task-specific settings","protocol_id":"1975ca8c09f87d3fc7597c12","raw_value":52.6,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"52.6%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row14:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1","source_record_id":"official-0dd3bed5c7dd:table0:row14:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":52.6},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","model_id":"m-98ee7005cf45b532","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"fb4474bb9b423ef25754804c5a39561f","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic scientific research Terminal-Bench-Science 0.1 [1]; preserve page footnotes and task-specific settings","protocol_id":"88023c6718be09c275d82cfc","raw_value":52.6,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"52.6%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row2:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1 [1]","source_record_id":"anthropic-fable51:table0:row2:column1","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":52.6},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"03754527ade1d4e46c831eb614495a2b","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":55.8,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"55.8% 60.9% (Mythos 5.1)","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row4:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"anthropic-fable51:table0:row4:column1","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":55.8},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"a5d016c51f0b0d94dac22663791c4a6b","model_id":"m-98ee7005cf45b532","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":31.4,"research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"31.4%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column1","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":31.4},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"cd6feaf81208453ed0051fee8e3b0d5a","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1ae0eb143abd6ce0f4452caf7560b910","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1478.4,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1478.4},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc3d289adb33b8c99688faaf7ba620e4","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1524.5,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1524.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e6f07c08fda105325a4e56b265e4dee","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1648.98,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1648.98},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e97d86463820bac312fd30b55101aa18","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1576.5,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1576.5},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"406d135dd93885891c99ba06991cb897","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5866666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5866666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f92093c1907c5885752360049477027e","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6256666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6256666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f0111de6b33e329eaa3eeaaa4641d38","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6716666666666666,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6716666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"696f19e1d154ec2d1b2db394ee639f7c","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.647,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9a4dfb7c1343b38cab40824ed5734fe","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.40161290322580645,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40161290322580645},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"83b0fc3e69397802ea873c43c47bf4b4","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30186999109528045,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30186999109528045},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"178921cf9f9c7f9c087a0211aa9278a9","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30456852791878175,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30456852791878175},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fdb802226467d3d56e9382a3ce01d898","model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2983947119924457,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2983947119924457},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e6e73416fd7cc92cc470effb47f660ee","model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.522404059498186,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.522404059498186},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85eee94e671af163217f6d12bfe92bb2","model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5466103465672885,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5466103465672885},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1ca9212f0aaf8dddbf37551cb9054ac0","model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.57772425084896,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.57772425084896},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a60805eb1c3fc953f4fea1f68e7874b8","model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5532309746463979,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5532309746463979},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"fd74c3c653ed9c2b342d0c63af5e691b","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.214285714285714,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-fable-5-1-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.214285714285714},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4106b7534aad183df90358d5970dd5bf","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.29047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-fable-5-1-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29047619047619},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9f1bb911398047de48a7366176d8fef","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.433333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-fable-5-1:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.433333333333333},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11cc383a18b4477133b9e9c14ed102bc","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.404761904761905,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-fable-5-1-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404761904761905},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"99d728497760c11f804edf3db0397825","model_id":"m-98ee7005cf45b532","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.395238095238095,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-fable-5-1-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395238095238095},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"68804b57f59e67581192a59590da9ac2","model_id":"m-98ee7005cf45b532","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.228400505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228400505050505},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c94e44aab63389014eb297f3edb294d4","model_id":"m-98ee7005cf45b532","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.287730808080808,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287730808080808},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","model_id":"m-98ee7005cf45b532","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33798ef7a7c3ffb01af78ba5b1d28657","model_id":"m-98ee7005cf45b532","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.357316919191919,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.357316919191919},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"593be206f43d234ef58751d27294cbdb","model_id":"m-98ee7005cf45b532","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.302979545454545,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.302979545454545},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c34d650ba1f5e0acc54c8a81258b3432","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.374,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-fable-5-1-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.374},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4378319d00527ae31b6350edbba14eb8","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-fable-5-1-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6cd57d6545a81a30a933d08c69e8d902","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.601,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-fable-5-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.601},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a14dae8f93889186c94f8b10b3616ea5","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-fable-5-1-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e1705b2cbbf6a0599b5cc40acdf215f0","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.504,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-fable-5-1-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.504},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9fe6c39bd7d079943d38e0d419769501","model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.28,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f8af06120cd189abddc3df4bb6853a5","model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.268,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.268},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0cc822bb01d821119228649fdf1c1421","model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0360eda71cf68bdd28fa47569b7576ce","model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.268,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.268},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f26f9b52eed42aacf6b18eb6bf2d5746","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.404040404040404,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404040404040404},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a67bc8182927b5571727b1f4c339a20b","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.44949494949495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44949494949495},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4e65adb49bfb1b584e89b82226fe99ba","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.55050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.55050505050505},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0abacd4982a76a965ebbaf64a7ade16","model_id":"m-98ee7005cf45b532","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0250c982341a8269bd1b765d5b9abae","model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"fe1fff2df783257f50aa06b58289ecd6","model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.585,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.585},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"131042ad05651fc18226a15142c8b260","model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.6,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4bfa4d8bbc3d0ebd4db2d684b1640d37","model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.62,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.62},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33dd21c6f50a26db637899e1d5fd814b","model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1476.7,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1476.7},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"17fa36dd0f0085d094e089dcf3133c93","model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.23,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.23},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e5e766037c0671e80db2b99e1b118a8","model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1740.22,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1740.22},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22bae45f48a48bf109a74535736a1268","model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1644.08,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1644.08},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"891daa40ba328005069aa0e8ba5b50ef","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.880808080808081,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1-low:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.880808080808081},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff38753334463880209f9bb6bd37d739","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.885858585858586,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1-medium:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.885858585858586},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7a3b4581fbf27099371300fe1f7621b1","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.934343434343434,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1-xhigh:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.934343434343434},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e192525c6b2961fba46c9db39251f766","model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.906060606060606,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1-high:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.906060606060606},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2893ee1a5163d51c4e8c94b66d42184e","model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.389690721649485,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.389690721649485},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc6462c9d688ef35bcd7f5726bfaa53f","model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-medium:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-medium:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"23b5f80e4b445ac75a8f8c836738a5b8","model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.457731958762887,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-xhigh:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-xhigh:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.457731958762887},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b03af4be01d6c8ac01fea2c55b790d30","model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.430927835051546,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-high:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-high:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.430927835051546}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":533596,"numerator":44974135},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":500},"family":"FrontierMath","row_ids":["a450c91f279ff648cb2b208a74552ba0"],"score":71.43},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5700},"family":"GPQA Diamond","row_ids":["0539c349124105a728b2ffaeadc2dabd"],"score":96.61},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1650},"family":"Terminal-Bench-Science","row_ids":["2012b47ec868b52e455d3d8d66a1a48a"],"score":86.84},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":325},"family":"HLE","row_ids":["16b64e0ecfeb64ebd3bc40ba58f5f691"],"score":81.25},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1450},"family":"CritPt","row_ids":["2a25d966a3bebfa161181074ba7fbad0"],"score":85.29}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":96.61,"percentile_exact":{"denominator":59,"numerator":5700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","peer_count":39,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":84.28,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":496053558,"numerator":42907839551},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":2375},"family":"tau3-Banking","row_ids":["a5ec26f61468360fa82a5667cce554db"],"score":91.35},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7600},"family":"GDPval-AA","row_ids":["1bc96450387b9110981d343f4dfd20eb"],"score":93.83},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":262},"family":"Terminal-Bench","row_ids":["71f828c6aa1f518ef08599ca0a1b0570"],"score":87.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1400},"family":"AutomationBench","row_ids":["9a8fd23ce0ae510627402be3083ec808"],"score":73.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7300},"family":"AA-Briefcase","row_ids":["cab25ff85e887187f7f900705800483c"],"score":94.81},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"OSWorld","row_ids":["520b3d1c3a49b9ec75dd20571c69ae96"],"score":66.67},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2250},"family":"AA-AnalystAgent","row_ids":["172f756dc0c16d9248bcf51116651a4b"],"score":97.83}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","peer_count":82,"percentile":93.83,"percentile_exact":{"denominator":81,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":78,"percentile":94.81,"percentile_exact":{"denominator":77,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.1","peer_count":7,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"score":86.5,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":98,"numerator":5325},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4100},"family":"FrontierCode","row_ids":["468e0160c503684573c12f005950e313"],"score":83.67},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["a9df38e2550ef9f8d566441fba3e6c66"],"score":25.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.0","peer_count":50,"percentile":83.67,"percentile_exact":{"denominator":49,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"1.1","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4}],"score":54.34,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":102714,"numerator":8141825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3425},"family":"GDP.pdf","row_ids":["5d9e8383609a67f5871ab595963f24e2"],"score":90.13},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1500},"family":"SimpleQA Verified","row_ids":["add598a99258de2861924fc25c825623"],"score":88.24},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3150},"family":"AA-Omniscience","row_ids":["b829940da89808936844666d294a78aa","2cbd1e01e64dad121dca7aaf300f12be"],"score":59.43}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996}],"score":79.27,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":9,"numerator":800},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":700},"family":"Humanity Sixth Sense","row_ids":["2d80b1607b40fc4d214cb22e259d1221"],"score":77.78},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["a7fcafd577ada1f13084bdc5d25fc8ab"],"score":100.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","peer_count":28,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4}],"score":88.89,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":96.61,"percentile_exact":{"denominator":59,"numerator":5700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","peer_count":39,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","peer_count":82,"percentile":93.83,"percentile_exact":{"denominator":81,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":78,"percentile":94.81,"percentile_exact":{"denominator":77,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.1","peer_count":7,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.0","peer_count":50,"percentile":83.67,"percentile_exact":{"denominator":49,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"1.1","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","peer_count":28,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4}],"result_updated_at":"2026-09-29","score":78.31,"score_exact":{"denominator":30136812673968,"numerator":2359877379520649}}],"release_date":"2026-09-01","release_status":"published","release_verified":true,"source_id":"anthropic/claude-fable-5-1"},"m-99183a5b3b302b0d":{"id":"m-99183a5b3b302b0d","in_window":true,"name":"DeepSeek V4.1 Flash","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"9641c48c69c00cab23b6294d207c8f62","model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":31.2,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"31.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":31.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94}],"official_evidence":{"artifact_sha256":"df2d95745d8bc7620e1be8267ede37334adde98086da78ba1e49574f407a619a","basis":"official_release_page","date":"2026-09-10","sha256":"df2d95745d8bc7620e1be8267ede37334adde98086da78ba1e49574f407a619a","url":"https://www.deepseek.com/en/news/deepseek-v4-1-flash/"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":513,"numerator":7325},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":50},"family":"FrontierMath","row_ids":["3121b8aa930fa12d352b2910814de607"],"score":5.56},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":100},"family":"GPQA Diamond","row_ids":["6c2c7071889205aa8d51f6bb7f06fb84"],"score":33.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["90672f02fa6c8adab2a2801b7c2aedb3"],"score":3.95}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"score":14.28,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":72,"numerator":5375},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":325},"family":"Terminal-Bench","row_ids":["4ce54d74c2045140751a930d5f81e63e"],"score":40.63},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["e28fafc89bbfaddac23f5b76027e9f84"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":250},"family":"AA-Briefcase","row_ids":["59421edaa49a1a1c8543f29553d821c6"],"score":83.33}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","peer_count":49,"percentile":40.63,"percentile_exact":{"denominator":8,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855}],"score":74.65,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":1100},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":900},"family":"Vals Code Migration","row_ids":["3beb1575e4852b44427673ea326ae656"],"score":69.23},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["3bd3c3d2983a2fb1db83fc1872674955"],"score":100.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":27,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2}],"score":84.62,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":550},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":550},"family":"SimpleQA Verified","row_ids":["66b5a46f01ac8af0a7586871cb050ba1"],"score":32.35}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429}],"score":32.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"Chartography","row_ids":["e5a64d88a17b8e69890219a69e61c957"],"score":33.33}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9}],"score":33.33,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","peer_count":49,"percentile":40.63,"percentile_exact":{"denominator":8,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":27,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9}],"result_updated_at":"2026-10-13","score":48.96,"score_exact":{"denominator":3627936,"numerator":177623315}}],"release_date":"2026-09-10","release_status":"published","release_verified":true,"source_id":"deepseek/deepseek-v4.1-flash"},"m-99ca714b8835cfb9":{"id":"m-99ca714b8835cfb9","in_window":true,"name":"Qwen3.8 Max","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"fb72cb8ea716369f0491dda0a2c1f8cc","model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":27.3,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"27.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":27.3},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"de593678281918b51aaf95adefe186dd","model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.6,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"9e8a69deb5a1faa9380e095362bd2e2c","model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":56.6,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"56.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":56.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"71646c5fb8bcced91b2a83b600993b46","model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":82.8,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"82.8","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":82.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"9c1d87da906e98d13d0eb89f034d7898","model_id":"m-99ca714b8835cfb9","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.2,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.2","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","model_id":"m-99ca714b8835cfb9","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"37b17217723693963901c38560492d58","model_id":"m-99ca714b8835cfb9","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":41.1,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.1},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6ff1592097b1d5537d18b6c552613651","model_id":"m-99ca714b8835cfb9","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":55.6,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"55.6","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":55.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86}],"official_evidence":{"artifact_sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","basis":"official_release_page","date":"2026-08-02","sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","url":"https://www.alibabacloud.com/help/en/model-studio/newly-released-models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":7075},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":175},"family":"GPQA Diamond","row_ids":["7ff900665f3a2edd6bd47eeb54dcc81e"],"score":58.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1250},"family":"Terminal-Bench-Science","row_ids":["a18d9b692129832a6701c8778a0404f3"],"score":65.79}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":58.33,"percentile_exact":{"denominator":3,"numerator":175},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","peer_count":39,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86}],"score":62.06,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":1008,"numerator":58825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":200},"family":"APEX-Agents","row_ids":["bf4bf643f4907b71184823fd0ebd6b1c"],"score":28.57},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":725},"family":"Terminal-Bench","row_ids":["e085a37b9093e9f4b5222ae848bc472f"],"score":60.42},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"AutomationBench","row_ids":["58a53dd6f007e225a58569d30c64eee8"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":850},"family":"AA-Briefcase","row_ids":["0d1d30147df15075a1ed15467a0ef1b4"],"score":94.44}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","included":true,"model_id":"m-99ca714b8835cfb9","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","peer_count":49,"percentile":60.42,"percentile_exact":{"denominator":12,"numerator":725},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"reference_revision":"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593}],"score":58.36,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["9619717553c9af1d9fdaa3ac01eac67f"],"score":25.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"reference_revision":"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6}],"score":25.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":17,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["57da33d0598ebe148085f216c7c7cefb"],"score":100.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":600},"family":"SimpleQA Verified","row_ids":["2e690ecc6ebeb390f087e351bd8c7803"],"score":35.29}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458}],"score":67.65,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":54,"numerator":3725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1700},"family":"Humanity Sixth Sense","row_ids":["b0de0689b04f1b90ca42e94e49e0a3bb"],"score":62.96},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":75},"family":"MMMU-Pro","row_ids":["d1f2e18c925df69cd229259da568a231"],"score":75.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","peer_count":28,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3}],"score":68.98,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":58.33,"percentile_exact":{"denominator":3,"numerator":175},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","peer_count":39,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","included":true,"model_id":"m-99ca714b8835cfb9","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","peer_count":49,"percentile":60.42,"percentile_exact":{"denominator":12,"numerator":725},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"reference_revision":"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"reference_revision":"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","peer_count":28,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3}],"result_updated_at":"2026-08-27","score":55.53,"score_exact":{"denominator":3907008,"numerator":216965705}}],"release_date":"2026-08-02","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.8-max"},"m-9e1e0f605d47cf02":{"id":"m-9e1e0f605d47cf02","in_window":true,"name":"Motif 3","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":2800},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2800},"family":"GPQA Diamond","row_ids":["8763ad1982a06ca4a70e322ffe1ddbd4"],"score":47.46}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":47.46,"percentile_exact":{"denominator":59,"numerator":2800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434}],"score":47.46,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":243243,"numerator":13098550},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":950},"family":"tau3-Banking","row_ids":["479061fd58fe09f756e592ab99fa1642"],"score":73.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3800},"family":"GDPval-AA","row_ids":["d1a5193baa7d56820cb5874aa60b7c4a"],"score":46.91},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3200},"family":"AA-Briefcase","row_ids":["56f3576f6b7e7bada3d4ea81a3e974d2"],"score":41.56}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","peer_count":82,"percentile":46.91,"percentile_exact":{"denominator":81,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","peer_count":78,"percentile":41.56,"percentile_exact":{"denominator":77,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75}],"score":53.85,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":500},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":500},"family":"GDP.pdf","row_ids":["b0db480f387627aef52b7f9b3d6804ac"],"score":26.32}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052}],"score":26.32,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":47.46,"percentile_exact":{"denominator":59,"numerator":2800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","peer_count":82,"percentile":46.91,"percentile_exact":{"denominator":81,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","peer_count":78,"percentile":41.56,"percentile_exact":{"denominator":77,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052}],"result_updated_at":null,"score":43.7,"score_exact":{"denominator":1908727821,"numerator":83411342375}}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"source_id":"motif-technologies/motif-3"},"m-9e460c49820cbe4e":{"id":"m-9e460c49820cbe4e","in_window":true,"name":"Kimi K3","observed_records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Kimi K3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7b257a4e69fd26a5efca3f99844cb10c","model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.2,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:Kimi K3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:Kimi K3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"f3a815210c20a49420a2227fcd8d8f19","model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":12.6,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"12.6","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":12.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"175facc5a97ac64a62324f8687068b9f","model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":67.5,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"67.5","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":67.5},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"ab9a351b7071ab7f587fbaf6f7ef4c51","model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":46.7,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"46.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":46.7},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"233e224a230e5689366760183b9c92e8","model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column K3. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":92.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"92.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":92.9},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"0949556ec5085216686632a53c4b2aa4","model_id":"m-9e460c49820cbe4e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":84.2,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"84.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":84.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Kimi K3","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"d4511580b7749fdce0eb519414c56c16","model_id":"m-9e460c49820cbe4e","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.8,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Kimi K3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.8},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Kimi K3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"742baf559c583daecd3dac4f871089d0","model_id":"m-9e460c49820cbe4e","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.2,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.2","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Kimi K3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.2},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Kimi K3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"9d73635ebbe34cac1967dc999b07ad7a","model_id":"m-9e460c49820cbe4e","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":45.5,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"45.5","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Kimi K3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":45.5},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Kimi K3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"9a67d273dfa56310e03c5e50dc85cc6e","model_id":"m-9e460c49820cbe4e","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":74.0,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"74.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Kimi K3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":74.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d90665c7d7b6aa499394e0fc04faecd3","model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5097236438075742,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"00f30fccbbab1ec345444ee8"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"00f30fccbbab1ec345444ee8","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5097236438075742},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59}],"official_evidence":{"artifact_sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","basis":"official_release_page","date":"2026-07-16","sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","url":"https://www.kimi.com/en/blog/"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":342,"numerator":12275},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":250},"family":"GPQA Diamond","row_ids":["dd3025c1130213f154e327d038efe0ed"],"score":83.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["f2b3352d05d0abb45473dcad237d2a14"],"score":11.84},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":25},"family":"HLE","row_ids":["0e05cbe602caeb64b25d2230f75c2113"],"score":12.5}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","peer_count":17,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2}],"score":35.89,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":72,"numerator":3887},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":80},"family":"APEX-Agents","row_ids":["e4b1483bcb889a5ef567c4c02fc571dd"],"score":80.0},{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":275},"family":"Terminal-Bench","row_ids":["fff5c31ca1adab72967ff7c824d4653e"],"score":34.38},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["cf571d5463813d8daec30e8a8f31740d"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":100},"family":"AA-Briefcase","row_ids":["04f715495ca6c55181e1b42a64312d78"],"score":11.11},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":400},"family":"MCP Atlas","row_ids":["b1436c94b33fa58416664a5d31b199a9"],"score":44.44}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3}],"score":53.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":450},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":250},"family":"Vals Code Migration","row_ids":["549f9102d89b3e9a3355e2ef41ef0184"],"score":19.23},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"DeepSWE","row_ids":["c87699e8afdc2f67ab86923665dd9ed5"],"score":50.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","peer_count":27,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5}],"score":34.62,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":50},"family":"SimpleQA Verified","row_ids":["63b449d94d8f27d0e405ea5ec8384866"],"score":50.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506}],"score":50.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":81,"numerator":2600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":800},"family":"Humanity Sixth Sense","row_ids":["d100a26c5dab08a99d576ce25ec3994e"],"score":29.63},{"benchmark_version":"Standard 10","exact":{"denominator":3,"numerator":200},"family":"MMMU-Pro","row_ids":["9ea024ab5ca2b01fdc71c3f50f184c9c"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["24019ec9f1eb6cf5c33ae9562d1f1bac"],"score":0.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","peer_count":28,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1}],"score":32.1,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","peer_count":17,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","peer_count":27,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","peer_count":28,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1}],"result_updated_at":"2026-08-27","score":42.6,"score_exact":{"denominator":640224,"numerator":27275101}}],"release_date":"2026-07-16","release_status":"published","release_verified":true,"source_id":"moonshotai/kimi-k3"},"m-9ff6170077842b96":{"id":"m-9ff6170077842b96","in_window":true,"name":"Nemotron 3.5 Lightning","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","model_id":"m-9ff6170077842b96","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","model_id":"m-9ff6170077842b96","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","model_id":"m-9ff6170077842b96","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","model_id":"m-9ff6170077842b96","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","model_id":"m-9ff6170077842b96","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","model_id":"m-9ff6170077842b96","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","model_id":"m-9ff6170077842b96","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","model_id":"m-9ff6170077842b96","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","model_id":"m-9ff6170077842b96","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":52678025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1600},"family":"GPQA Diamond","row_ids":["02699c8c436679b49b6319aa832d4d14"],"score":27.12},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["ac28090aae70acdb086bd22103c739d0"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":500},"family":"HLE","row_ids":["e12a9fbbd9fdb2bbd3147c3b6679efcd"],"score":7.94},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["2f96f5af894d02c94d57177465def23b"],"score":11.76}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":27.12,"percentile_exact":{"denominator":59,"numerator":1600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond","peer_count":64,"percentile":7.94,"percentile_exact":{"denominator":63,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":16.03,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":22949483},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":200},"family":"tau3-Banking","row_ids":["612f78abda8cf62d29f104d80c127636"],"score":15.38},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":500},"family":"GDPval-AA","row_ids":["1694dd426967f516c30dcdcc5616df9b"],"score":18.52},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["803b4920eccad03514634ebd3575a2d8"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":150},"family":"AutomationBench","row_ids":["ec3633dbe4490f5d8334e3da32d471af"],"score":7.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1200},"family":"AA-Briefcase","row_ids":["47de14e8ef5807902d31f4e62879bab3"],"score":15.58},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"EnterpriseOps-Gym","row_ids":["80ae66f2b7393fe1f51bab7846dd0a10"],"score":0.0}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2.1","peer_count":82,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":78,"percentile":15.58,"percentile_exact":{"denominator":77,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA run","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537}],"score":14.9,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["c503c45f721e54502bd0c3c277c347cb"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":46875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":275},"family":"GDP.pdf","row_ids":["40a272cc2b1879bb5b0104ebb8833d18"],"score":14.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1700},"family":"AA-Omniscience","row_ids":["8698ac189bdac0fe4a6655a8ed1ade73","2184077575074c69e38c71abae562f14"],"score":32.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636}],"score":23.27,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":27.12,"percentile_exact":{"denominator":59,"numerator":1600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond","peer_count":64,"percentile":7.94,"percentile_exact":{"denominator":63,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2.1","peer_count":82,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":78,"percentile":15.58,"percentile_exact":{"denominator":77,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA run","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636}],"result_updated_at":null,"score":14.44,"score_exact":{"denominator":41274330401304,"numerator":596156995131155}}],"release_date":"2026-08-11","release_status":"published","release_verified":true,"source_id":"nvidia/nemotron-3.5-lightning"},"m-a3a34c5608c68d32":{"id":"m-a3a34c5608c68d32","in_window":true,"name":"Qwen3.7 Max","observed_records":[{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","model_id":"m-a3a34c5608c68d32","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"abbe21aa88bf4a572f15ca2233e791d6","model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":21.6,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"21.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":21.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"ffb82b85f11b529dafa9689c16b28131","model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":14.2,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"14.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":14.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"68966a1d85f137636ad7fa5c27b6e2d8","model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":79.1,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"79.1","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":79.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"a6934af3407d37537eb6eb1bc8d0e969","model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.4,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.4","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.4},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"official_evidence":{"artifact_sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","basis":"official_release_page","date":"2026-05-21","sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","url":"https://www.alibabacloud.com/help/en/model-studio/newly-released-models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1,"numerator":40},"families":[{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":40},"family":"GPQA Diamond","row_ids":["8afb3ab2b09d75bd6e3e04698e2c3cb8"],"score":40.0}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"reference_revision":"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4}],"score":40.0,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":18,"numerator":925},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["4ea5a3b77e03d03f1422a6728fc0742b"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":50},"family":"AA-Briefcase","row_ids":["680700051014bfbbbe521e958802f31c"],"score":5.56},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["0886e98ce0984c34f09c65e2184f8692"],"score":100.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":100},"family":"MCP Atlas","row_ids":["f29e77fd200946b25fe37a0df78120da"],"score":100.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"reference_revision":"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4}],"score":51.39,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["2206795ef677e458e6f834638793b28d"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"reference_revision":"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":34,"numerator":1945},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":70},"family":"IFBench","row_ids":["1be1bd504f0391002f7de4a46514d7c5"],"score":70.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"IFEval","row_ids":["98b4a1e95c1679056cdd34a216b7399d"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"MMLU-Pro","row_ids":["3e7111ab4e0b4463d98c531d78381779"],"score":100.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1000},"family":"SimpleQA Verified","row_ids":["cb8036f627af27545731c0fdb0405c49"],"score":58.82}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","peer_count":6,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"score":57.21,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"reference_revision":"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"reference_revision":"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"reference_revision":"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","peer_count":6,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"result_updated_at":"2026-08-27","score":38.1,"score_exact":{"denominator":5508,"numerator":209845}}],"release_date":"2026-05-21","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.7-max"},"m-a3f1f1de993a039e":{"id":"m-a3f1f1de993a039e","in_window":true,"name":"Ling 3.1 Flash 1019","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":5967,"numerator":358000},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["6b7d4617c648ad490ad35eb982d8d2e7"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":500},"family":"HLE","row_ids":["39803a6fda08687f289513743c809091"],"score":55.56},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1200},"family":"CritPt","row_ids":["20d35e57ebd245b7509c38232fe59998"],"score":70.59}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","peer_count":64,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","peer_count":52,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385}],"score":60.0,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":158004,"numerator":12977233},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2300},"family":"GDPval-AA","row_ids":["f04f5a2ca29d7ef8a19efd2486d1e259"],"score":85.19},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":224},"family":"Terminal-Bench","row_ids":["cf0c6cf90a55894b4c003eacdea2ccb7"],"score":74.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1675},"family":"AutomationBench","row_ids":["115a839b24d533001f549fec19f24cc9"],"score":88.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6200},"family":"AA-Briefcase","row_ids":["2d7e1c988db8f9f1d1a2378f0202de51"],"score":80.52}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","peer_count":82,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":78,"percentile":80.52,"percentile_exact":{"denominator":77,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522}],"score":82.13,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":2300},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2300},"family":"FrontierCode","row_ids":["a3fd831de30519528e597eb5e2ceea3e"],"score":46.94}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","peer_count":50,"percentile":46.94,"percentile_exact":{"denominator":49,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151}],"score":46.94,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":205325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2225},"family":"GDP.pdf","row_ids":["d7a0ab07d52625e0d345f1bbf275edda"],"score":58.55},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["568a15734298d304532a5f2a4188d632","ae54534276a6c4c19dcb828d96fdee4c"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247}],"score":50.97,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","peer_count":64,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","peer_count":52,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","peer_count":82,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":78,"percentile":80.52,"percentile_exact":{"denominator":77,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","peer_count":50,"percentile":46.94,"percentile_exact":{"denominator":49,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247}],"result_updated_at":null,"score":61.24,"score_exact":{"denominator":233188307352,"numerator":14280161662615}}],"release_date":"2026-10-22","release_status":"published","release_verified":true,"source_id":"published/aa:ling-3-1-flash-rc5"},"m-a4be17f49ae915ce":{"id":"m-a4be17f49ae915ce","in_window":true,"name":"Gemma 4 12B","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"46666104ebdf3991c407d7606f028e64","model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.451700680272109,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b-non-reasoning:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.451700680272109},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6c83e52b0c0a02fd2b566c423daffb51","model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.619653179190751,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.619653179190751},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"aff3f1f1a247b8c1bfe9024113fc4696","model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.660606060606061,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.660606060606061},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":1850},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1850},"family":"GPQA Diamond","row_ids":["6fa185e6f3a52003678d56e35ea14de0"],"score":31.36}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252}],"score":31.36,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1539,"numerator":10700},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":700},"family":"GDPval-AA","row_ids":["60ed1cde923cee15c4935672b177aa62"],"score":8.64},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":100},"family":"AutomationBench","row_ids":["49cf1a20c5edfdf8e5e8bfd178db37ad"],"score":5.26}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","peer_count":82,"percentile":8.64,"percentile_exact":{"denominator":81,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258}],"score":6.95,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4,"numerator":325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":325},"family":"IFBench","row_ids":["af2b7a8fe076debb535a31ef7b09244c"],"score":81.25}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864}],"score":81.25,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":250},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":250},"family":"MMMU-Pro","row_ids":["ca0bc4ca214d2607b60ef7f92f73e72e"],"score":14.71}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514}],"score":14.71,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","peer_count":82,"percentile":8.64,"percentile_exact":{"denominator":81,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514}],"result_updated_at":null,"score":34.12,"score_exact":{"denominator":24697872,"numerator":842744275}}],"release_date":"2026-06-03","release_status":"published","release_verified":true,"source_id":"published/aa:gemma-4-12b"},"m-a67d9ce95a33f246":{"id":"m-a67d9ce95a33f246","in_window":true,"name":"K2 Horizon MoVA 36B A4B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":117351,"numerator":3265325},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2400},"family":"GPQA Diamond","row_ids":["da90bd5db2e44d989ddf563c8ec38c6b"],"score":40.68},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["13146daedf685d8da2f4bf015457ee2c"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1300},"family":"CritPt","row_ids":["d6e7cb95527cd29a8925ca94f594e568"],"score":25.49}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":40.68,"percentile_exact":{"denominator":59,"numerator":2400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","peer_count":52,"percentile":25.49,"percentile_exact":{"denominator":51,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08}],"score":27.83,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":24453,"numerator":1164503},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":875},"family":"tau3-Banking","row_ids":["7e3c134e84deebeefe2308a65aef1578"],"score":67.31},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":500},"family":"GDPval-AA","row_ids":["7f3c9ed4efe6520cbbf9a35bd5a39f5e"],"score":55.56},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["cd7e16f694d6c0543ca83b3b5dd9358b"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":900},"family":"AutomationBench","row_ids":["908bb98659898bf60c101cff97ce69b8"],"score":47.37},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":600},"family":"AA-Briefcase","row_ids":["54e0dbe28b524361d2c39e892adaa185"],"score":54.55}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":67.31,"percentile_exact":{"denominator":13,"numerator":875},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","peer_count":82,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":47.37,"percentile_exact":{"denominator":19,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","peer_count":78,"percentile":54.55,"percentile_exact":{"denominator":11,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79}],"score":47.62,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1200},"family":"FrontierCode","row_ids":["33a3261e51180ab895a82ccd6f7e12a3"],"score":24.49}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","peer_count":50,"percentile":24.49,"percentile_exact":{"denominator":49,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444}],"score":24.49,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":825},"family":"GDP.pdf","row_ids":["14efb36fead599ff0b956fa101f23a86"],"score":21.71}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042}],"score":21.71,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":40.68,"percentile_exact":{"denominator":59,"numerator":2400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","peer_count":52,"percentile":25.49,"percentile_exact":{"denominator":51,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":67.31,"percentile_exact":{"denominator":13,"numerator":875},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","peer_count":82,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":47.37,"percentile_exact":{"denominator":19,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","peer_count":78,"percentile":54.55,"percentile_exact":{"denominator":11,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","peer_count":50,"percentile":24.49,"percentile_exact":{"denominator":49,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042}],"result_updated_at":null,"score":31.22,"score_exact":{"denominator":3605374773,"numerator":112575457730}}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"source_id":"published/aa:k2-horizon-mova-36b-a4b"},"m-ad7decc9ae94df12":{"id":"m-ad7decc9ae94df12","in_window":true,"name":"Nemotron 3 Ultra 550B A55B","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-04","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":15,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1095276,"numerator":37020775},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3700},"family":"GPQA Diamond","row_ids":["198f62e91fe9f1a3e9e87c6a5a19a9fe"],"score":62.71},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["70737e236c952ff057602f3190f37b75"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":500},"family":"HLE","row_ids":["85bd1df1c636ca5159d1bed6e1a0f499"],"score":23.81},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1600},"family":"CritPt","row_ids":["f5fbcfce1c679af1818579e2368d3d27"],"score":31.37}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":62.71,"percentile_exact":{"denominator":59,"numerator":3700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","peer_count":64,"percentile":23.81,"percentile_exact":{"denominator":21,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","peer_count":52,"percentile":31.37,"percentile_exact":{"denominator":51,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15}],"score":33.8,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":10783773,"numerator":221054348},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":300},"family":"tau3-Banking","row_ids":["5c824422e02d6ae8e8e0e7805866dd38"],"score":23.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2800},"family":"GDPval-AA","row_ids":["d9fac2d0e8fa2fae504adca322b42dfa"],"score":34.57},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["22a0f772949251594cd3a25e02ef9752"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":350},"family":"AutomationBench","row_ids":["b55d26365439e3d2ed6eba908f339027"],"score":18.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2300},"family":"AA-Briefcase","row_ids":["e12061e69225a05e579dff8f6ab0f0de"],"score":29.87},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["4c1c120149f7031df34c6f3ac8ff9629"],"score":5.56},{"benchmark_version":"current AA pass^5","exact":{"denominator":1,"numerator":0},"family":"AA-AnalystAgent","row_ids":["2275301a736d0fe2b34ce5c990945f54"],"score":0.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","peer_count":82,"percentile":34.57,"percentile_exact":{"denominator":81,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":78,"percentile":29.87,"percentile_exact":{"denominator":77,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","peer_count":24,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625}],"score":20.5,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":7,"numerator":100},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":100},"family":"FrontierCode","row_ids":["07260be02cf7c09201c1af77d371a386"],"score":14.29}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","peer_count":50,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576}],"score":14.29,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":232025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":375},"family":"IFBench","row_ids":["74e2baf55ca27065ec7aeb8f6eee9414"],"score":93.75},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":925},"family":"GDP.pdf","row_ids":["994833badab5af26add99f20244353b3"],"score":24.34},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2900},"family":"AA-Omniscience","row_ids":["7a931d4238635511e69e1feb35bfd055","f659a47d7f8491b1073ef450bc951aa0"],"score":54.72}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","peer_count":17,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645}],"score":57.6,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":62.71,"percentile_exact":{"denominator":59,"numerator":3700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","peer_count":64,"percentile":23.81,"percentile_exact":{"denominator":21,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","peer_count":52,"percentile":31.37,"percentile_exact":{"denominator":51,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","peer_count":82,"percentile":34.57,"percentile_exact":{"denominator":81,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":78,"percentile":29.87,"percentile_exact":{"denominator":77,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","peer_count":24,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","peer_count":50,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","peer_count":17,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645}],"result_updated_at":null,"score":31.06,"score_exact":{"denominator":41274330401304,"numerator":1281914164344215}}],"release_date":"2026-06-04","release_status":"published","release_verified":true,"source_id":"nvidia/nemotron-3-ultra-550b-a55b"},"m-af0edba2f76c80d6":{"id":"m-af0edba2f76c80d6","in_window":true,"name":"G9v3-39A5B","observed_records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-20","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":3,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":50.0,"evaluator_count":1,"exact":{"denominator":59,"numerator":2300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2300},"family":"GPQA Diamond","row_ids":["29813bdc52dc98722342b740c2f2d3f2"],"score":38.98}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":38.98,"percentile_exact":{"denominator":59,"numerator":2300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505}],"score":38.98,"weight":25},{"effective_weight_percent":50.0,"evaluator_count":1,"exact":{"denominator":234,"numerator":11275},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":675},"family":"tau3-Banking","row_ids":["76481bf229eeeddb8b44a15648197449"],"score":51.92},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":400},"family":"GDPval-AA","row_ids":["5d726bb21290d73d92a00bf9993c5a4a"],"score":44.44}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":51.92,"percentile_exact":{"denominator":13,"numerator":675},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","peer_count":82,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37}],"score":48.18,"weight":25}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":38.98,"percentile_exact":{"denominator":59,"numerator":2300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":51.92,"percentile_exact":{"denominator":13,"numerator":675},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","peer_count":82,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37}],"result_updated_at":null,"score":43.58,"score_exact":{"denominator":27612,"numerator":1203425}}],"release_date":"2026-08-20","release_status":"published","release_verified":true,"source_id":"published/aa:g9v3-39a5b"},"m-b0c41bb29e51489d":{"id":"m-b0c41bb29e51489d","in_window":true,"name":"Grok 4.5","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59}],"official_evidence":{"artifact_sha256":"605d18dafa601ac99d2ffe30c675010636f4c489483187f70c548f59080af7bf","basis":"official_release_page","date":"2026-07-16","sha256":"605d18dafa601ac99d2ffe30c675010636f4c489483187f70c548f59080af7bf","url":"https://x.ai/news/grok-4-5"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":38.46,"evaluator_count":2,"exact":{"denominator":24,"numerator":1025},"families":[{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":75},"family":"Terminal-Bench","row_ids":["a49b610cdf02973ee5519a3d58738257"],"score":18.75},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":200},"family":"AA-Briefcase","row_ids":["99d362c58056009c0378c89776bd8bfa"],"score":66.67}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","peer_count":49,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071}],"score":42.71,"weight":25},{"effective_weight_percent":30.77,"evaluator_count":2,"exact":{"denominator":4,"numerator":175},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["f4858ed8f9e084fde242d77a18dddeed"],"score":62.5},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":25},"family":"MirrorCode","row_ids":["09300d4eee5c7d673f52949c9b3dcb98"],"score":25.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224}],"score":43.75,"weight":20},{"effective_weight_percent":30.77,"evaluator_count":1,"exact":{"denominator":17,"numerator":750},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":750},"family":"SimpleQA Verified","row_ids":["5fb5dec681fa13f63b1b4d2e3c76490c"],"score":44.12}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483}],"score":44.12,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","peer_count":49,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483}],"result_updated_at":"2026-10-21","score":43.46,"score_exact":{"denominator":5304,"numerator":230525}}],"release_date":"2026-07-16","release_status":"published","release_verified":true,"source_id":"xai/grok-4.5"},"m-b439c804161fac26":{"id":"m-b439c804161fac26","in_window":true,"name":"Inkling","observed_records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"inkling","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"6e97ec16100fc8b44c1121b2536cf94d","model_id":"m-b439c804161fac26","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":11.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c3b7eadef9e2b3a75f1d7f1c"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"c3b7eadef9e2b3a75f1d7f1c","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":11.2},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"ea8212294d42fc5eee3233e729cf83c0","model_id":"m-b439c804161fac26","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.51,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.51%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row54"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row54","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.51},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","model_id":"m-b439c804161fac26","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","model_id":"m-b439c804161fac26","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","model_id":"m-b439c804161fac26","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","model_id":"m-b439c804161fac26","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","model_id":"m-b439c804161fac26","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","model_id":"m-b439c804161fac26","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","model_id":"m-b439c804161fac26","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","model_id":"m-b439c804161fac26","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","model_id":"m-b439c804161fac26","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","model_id":"m-b439c804161fac26","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":123000875},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3800},"family":"GPQA Diamond","row_ids":["a95cdd62235bf41ff0b78b6a5a866d88"],"score":64.41},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["8120a268fd2b22fe2a65df299e1ab0c5"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2000},"family":"HLE","row_ids":["2d291e3811085b2648de857339509203"],"score":31.75},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1850},"family":"CritPt","row_ids":["a9467dab432c3fd1ab9b333380c66f83"],"score":36.27}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":64.41,"percentile_exact":{"denominator":59,"numerator":3800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","included":true,"model_id":"m-b439c804161fac26","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond","peer_count":64,"percentile":31.75,"percentile_exact":{"denominator":63,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18}],"score":37.43,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":70864794,"numerator":2278099781},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":850},"family":"tau3-Banking","row_ids":["22b4395fa3e0e77e06e9fcfe630f8789"],"score":65.38},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3200},"family":"GDPval-AA","row_ids":["ff011af995e95f243ad573421cc4ba05"],"score":39.51},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["1344e9967afa11034830d610f592ed83"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":450},"family":"AutomationBench","row_ids":["5bd1e6ef9a7eb6f30e6c6bf694074f2d"],"score":23.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2000},"family":"AA-Briefcase","row_ids":["7d5fcb038f359de5556a57151b51cffc"],"score":25.97},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":350},"family":"EnterpriseOps-Gym","row_ids":["40747d0e01fbdeb26f3ad425860f13bb"],"score":38.89},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["73ec2c70e1808cc25e2b747786aa5d99"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":500},"family":"AA-AnalystAgent","row_ids":["753dbaee803264fefa51f263a84b1509"],"score":21.74}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","included":true,"model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2.1","peer_count":82,"percentile":39.51,"percentile_exact":{"denominator":81,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":23.68,"percentile_exact":{"denominator":19,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":78,"percentile":25.97,"percentile_exact":{"denominator":77,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA run","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","peer_count":24,"percentile":21.74,"percentile_exact":{"denominator":23,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375}],"score":32.15,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":49,"numerator":550},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1100},"family":"FrontierCode","row_ids":["9b48c9ab902cc3f53ca099a627db7e76"],"score":22.45},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"Vals Code Migration","row_ids":["2bad0033060ca0f870378d7a7d3d6c5d"],"score":0.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","included":true,"model_id":"m-b439c804161fac26","original_version_label":"2.0","peer_count":50,"percentile":22.45,"percentile_exact":{"denominator":49,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1.1","peer_count":27,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82}],"score":11.22,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2115350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1150},"family":"GDP.pdf","row_ids":["af7da6ed00064e0d491020c30636379c"],"score":60.53},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":350},"family":"SimpleQA Verified","row_ids":["28411df2fb9e1660a30498ca8b0c9245"],"score":20.59},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["95037c24ac83fd0eadfbf39605ac569f","3698e24fd77dd4f8bc27edd44ed2efc7"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","included":true,"model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329}],"score":41.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":350},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":350},"family":"MMMU-Pro","row_ids":["874b32395a8c924c1d3f0faf3d2e2b2e"],"score":20.59}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856}],"score":20.59,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":64.41,"percentile_exact":{"denominator":59,"numerator":3800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","included":true,"model_id":"m-b439c804161fac26","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond","peer_count":64,"percentile":31.75,"percentile_exact":{"denominator":63,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","included":true,"model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2.1","peer_count":82,"percentile":39.51,"percentile_exact":{"denominator":81,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":23.68,"percentile_exact":{"denominator":19,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":78,"percentile":25.97,"percentile_exact":{"denominator":77,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA run","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","peer_count":24,"percentile":21.74,"percentile_exact":{"denominator":23,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","included":true,"model_id":"m-b439c804161fac26","original_version_label":"2.0","peer_count":50,"percentile":22.45,"percentile_exact":{"denominator":49,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1.1","peer_count":27,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","included":true,"model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856}],"result_updated_at":"2026-08-27","score":29.94,"score_exact":{"denominator":16227514516752,"numerator":485799335327597}}],"release_date":"2026-07-15","release_status":"published","release_verified":true,"source_id":"thinkingmachines/inkling"},"m-b6a574cc5300d5f1":{"id":"m-b6a574cc5300d5f1","in_window":true,"name":"Beam 1.0","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"82df891de38afd61fe2c16983a26faa8","model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.04,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"4.04%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row49"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row49","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":4.04},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c7d3f53d83f5e19ff46dc9c5f753d703","model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"reflection-beam:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-16","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":61047,"numerator":2006075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["04366109cc25918e59cfe4140983a31b"],"score":11.84},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2500},"family":"HLE","row_ids":["8319710b72494f93151d3337f0dbcdfd"],"score":39.68},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":800},"family":"CritPt","row_ids":["bd76f518c9968d380eda9d90a689c8c2"],"score":47.06}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","peer_count":64,"percentile":39.68,"percentile_exact":{"denominator":63,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","peer_count":52,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265}],"score":32.86,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":158004,"numerator":7953793},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1400},"family":"GDPval-AA","row_ids":["522de4d527da35b85f7ead36d5bab4a3"],"score":51.85},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":154},"family":"Terminal-Bench","row_ids":["7a6b6215e85870e72209d73fd4101db4"],"score":51.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1125},"family":"AutomationBench","row_ids":["981d2354449aefd49dcd7a2e96840b24"],"score":59.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3000},"family":"AA-Briefcase","row_ids":["fb238aac3df6d482092b3981302c93d1"],"score":38.96}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","peer_count":82,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":59.21,"percentile_exact":{"denominator":19,"numerator":1125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":78,"percentile":38.96,"percentile_exact":{"denominator":77,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2}],"score":50.34,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":91,"numerator":1825},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":200},"family":"FrontierCode","row_ids":["06196596577df812dcdeed161680d16c"],"score":28.57},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":150},"family":"Vals Code Migration","row_ids":["baba3c8fb65a67f1bc8fc2ae1572a4d3"],"score":11.54}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","peer_count":50,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","peer_count":27,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85}],"score":20.05,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":98025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1025},"family":"GDP.pdf","row_ids":["25ac15cd539fc43d4a096d3dea9f7966"],"score":53.95},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["11dc653c29e2dbe67eaa7a540042fef3","111f3e2d8b49ec87b3bbe30f22c6238c"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239}],"score":48.67,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","peer_count":64,"percentile":39.68,"percentile_exact":{"denominator":63,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","peer_count":52,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","peer_count":82,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":59.21,"percentile_exact":{"denominator":19,"numerator":1125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":78,"percentile":38.96,"percentile_exact":{"denominator":77,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","peer_count":50,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","peer_count":27,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239}],"result_updated_at":null,"score":38.38,"score_exact":{"denominator":137088952,"numerator":5261997115}}],"release_date":"2026-10-16","release_status":"published","release_verified":true,"source_id":"published/aa:reflection-beam"},"m-b7faaf31060e95f7":{"id":"m-b7faaf31060e95f7","in_window":true,"name":"Solar Open2 250B","observed_records":[{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA-Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"24b5b69a17fb0e9c1efc24d9973f4c36","model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA-Diamond; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"62f4969756e5cb55e0903043","raw_value":86.3,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.3","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row3:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA-Diamond","source_record_id":"twelve-solar-card:table1:row3:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.3},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench (v6)","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"coding","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"2f29b0333ab4475d5bf842ed6cbb3b1f","model_id":"m-b7faaf31060e95f7","original_version_label":"release_v6; window and repeat count kept in protocol","profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench (v6); exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"3ac5a9c4884baf938debbd5e","raw_value":92.4,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"92.4","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench (v6)","source_record_id":"twelve-solar-card:table1:row5:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":92.4},{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"1322aa6b47c4d5bd57b13c62ff908257","model_id":"m-b7faaf31060e95f7","original_version_label":"2026","profile_key":"native","protocol_description":"Official published comparison; AIME2026; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"9f72a336f7b2466bee455f5d","raw_value":95.7,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"95.7","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row8:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME2026","source_record_id":"twelve-solar-card:table1:row8:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":95.7},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"Multi-Challenge","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"db8370e949230c30afaa0af7f9739bf4","model_id":"m-b7faaf31060e95f7","original_version_label":"Current published dataset; judge/configuration preserved","profile_key":"native","protocol_description":"Official published comparison; Multi-Challenge; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"c7ad72f37bc21288559c3655","raw_value":61.0,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"61.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row10:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MultiChallenge:Multi-Challenge","source_record_id":"twelve-solar-card:table1:row10:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":61.0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["f0716aa67008d67c9923512a4d8c5e11"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":162,"numerator":3275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["01aeec9c08a057635f9b72825df4f3fb"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":50},"family":"tau3-Banking","row_ids":["e24b9f9c121b7281ff7f37b4b404f354"],"score":50.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2500},"family":"GDPval-AA","row_ids":["7bf77397fcea6af7a97a2524f4c194bb"],"score":30.86},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["b96a70ae7b2ded290b5f169ceedf1e03"],"score":0.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","peer_count":82,"percentile":30.86,"percentile_exact":{"denominator":81,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2}],"score":20.22,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":9,"numerator":850},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"IFBench","row_ids":["dafb6731eb5f8725711540078af7013b"],"score":88.89},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"MMLU-Pro","row_ids":["27d4cca7113cd0216eb12449ea9867c2"],"score":100.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2}],"score":94.44,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","peer_count":82,"percentile":30.86,"percentile_exact":{"denominator":81,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2}],"result_updated_at":null,"score":54.18,"score_exact":{"denominator":133812,"numerator":7249925}}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"source_id":"published/aa:solar-open2-250b"},"m-b88efd1e16a0b031":{"id":"m-b88efd1e16a0b031","in_window":true,"name":"Tencent Hy4","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39}],"official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-11-03","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":27846,"numerator":2129875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1975},"family":"Terminal-Bench-Science","row_ids":["bb1f3609ad48f7662af1e4f3a15cef7a"],"score":75.96},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1700},"family":"HLE","row_ids":["5198f9fefccca8b5961dab3ccd43cfb2"],"score":80.95},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3700},"family":"CritPt","row_ids":["6c2b0cbfd16faf311bd4edca68e40d0e"],"score":72.55}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","peer_count":64,"percentile":80.95,"percentile_exact":{"denominator":21,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","peer_count":52,"percentile":72.55,"percentile_exact":{"denominator":51,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39}],"score":76.49,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1197,"numerator":95609},"families":[{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":248},"family":"Terminal-Bench","row_ids":["b9ebef353db3bd587384633039b7a5bb"],"score":82.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1625},"family":"AutomationBench","row_ids":["01aaf3c58f06d7a8962fe3fd60fa4286"],"score":85.53},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":500},"family":"AA-Briefcase","row_ids":["732f112b308243fb337093d80efff0c3"],"score":71.43}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","peer_count":76,"percentile":82.67,"percentile_exact":{"denominator":3,"numerator":248},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":85.53,"percentile_exact":{"denominator":19,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":78,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69}],"score":79.87,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":4000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4000},"family":"FrontierCode","row_ids":["8ed6fd9401411cecf6b0e548f78e12a5"],"score":81.63}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","peer_count":50,"percentile":81.63,"percentile_exact":{"denominator":49,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101}],"score":81.63,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1007,"numerator":64350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1550},"family":"GDP.pdf","row_ids":["b84d512b0d7ca681d5568eba0c2a9675"],"score":81.58},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2450},"family":"AA-Omniscience","row_ids":["795265b8e61a4f96a23b56d372ad2050","7d74696c1e12470b21a043973afd962c"],"score":46.23}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356}],"score":63.9,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","peer_count":64,"percentile":80.95,"percentile_exact":{"denominator":21,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","peer_count":52,"percentile":72.55,"percentile_exact":{"denominator":51,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","peer_count":76,"percentile":82.67,"percentile_exact":{"denominator":3,"numerator":248},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":85.53,"percentile_exact":{"denominator":19,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":78,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","peer_count":50,"percentile":81.63,"percentile_exact":{"denominator":49,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356}],"result_updated_at":null,"score":75.77,"score_exact":{"denominator":392572908,"numerator":29747186485}}],"release_date":"2026-11-03","release_status":"published","release_verified":true,"source_id":"published/aa:hunyuan-dev1030-aa"},"m-b9d9f5989f59e750":{"id":"m-b9d9f5989f59e750","in_window":true,"name":"Mercury 2.5","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3f9c90b35f6b5c160dd7d713ceef5f66","model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row57"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row57","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-08","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":2394,"numerator":20875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["80525908530cba67da8858ed258e1742"],"score":3.95},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":850},"family":"HLE","row_ids":["c6b63609fd0ed9a9d1cc3300620f0885"],"score":13.49}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051}],"score":8.72,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1881,"numerator":23195},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":100},"family":"GDPval-AA","row_ids":["21dd109a3c14f1e9b389fa0a30a5ee0f"],"score":11.11},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["179a39191c3950939aafcb199439b020"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":300},"family":"AutomationBench","row_ids":["43181abb2675b0ef85c7558935e2c229"],"score":15.79},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":100},"family":"AA-Briefcase","row_ids":["c4bef304832e9ad8f150cf2ae6ad2c77"],"score":9.09}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","peer_count":82,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":15.79,"percentile_exact":{"denominator":19,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","peer_count":78,"percentile":9.09,"percentile_exact":{"denominator":11,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49}],"score":12.33,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":350},"family":"GDP.pdf","row_ids":["8f8d77f644fefbdcb7d35712e527c726"],"score":18.42}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036}],"score":18.42,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","peer_count":82,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":15.79,"percentile_exact":{"denominator":19,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","peer_count":78,"percentile":9.09,"percentile_exact":{"denominator":11,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036}],"result_updated_at":null,"score":12.78,"score_exact":{"denominator":40964,"numerator":523575}}],"release_date":"2026-09-08","release_status":"published","release_verified":true,"source_id":"inception/mercury-2.5"},"m-bf7f21a45c57e0f7":{"id":"m-bf7f21a45c57e0f7","in_window":true,"name":"Muse Glimmer","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":96088925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2900},"family":"GPQA Diamond","row_ids":["ff2b5baedd439ff80001794125500940"],"score":49.15},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["f04dc7e5afab13c2812e16f8a7f09157"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1700},"family":"HLE","row_ids":["67c349c9897445ce219d7f8acc0e4dbe"],"score":26.98},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":400},"family":"CritPt","row_ids":["142184a428fc22f5f377c4fcd96f6a10"],"score":23.53}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":49.15,"percentile_exact":{"denominator":59,"numerator":2900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","peer_count":64,"percentile":26.98,"percentile_exact":{"denominator":63,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","peer_count":52,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07}],"score":29.24,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":22547889,"numerator":644722264},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":750},"family":"tau3-Banking","row_ids":["4a499df1f84b1eaf0a172cfe97a95f4c"],"score":57.69},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2000},"family":"GDPval-AA","row_ids":["cab0e8633e6425575d15983329eb67d3"],"score":24.69},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["af8dab2781940ab31c2449f02ac9e502"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":500},"family":"AutomationBench","row_ids":["af88768655cfd944c0cc2a22a6f5661c"],"score":26.32},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":100},"family":"AA-Briefcase","row_ids":["6b81d5a0603a102ca44c8276f19acede"],"score":14.29},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":250},"family":"EnterpriseOps-Gym","row_ids":["6751eebce76aacc484338eb2b0237376"],"score":27.78},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":400},"family":"AA-AnalystAgent","row_ids":["809f69850cfb92e35fdef7e6fcd0b405"],"score":17.39}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","peer_count":82,"percentile":24.69,"percentile_exact":{"denominator":81,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":78,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","peer_count":24,"percentile":17.39,"percentile_exact":{"denominator":23,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2}],"score":28.59,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":900},"family":"FrontierCode","row_ids":["eae24b4f084dd40f578e44715aebd484"],"score":18.37}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","peer_count":50,"percentile":18.37,"percentile_exact":{"denominator":49,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818}],"score":18.37,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":137425},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1625},"family":"GDP.pdf","row_ids":["0849343c6ca0b1de586bda128e33678b"],"score":42.76},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1350},"family":"AA-Omniscience","row_ids":["ce91d9d68242cca613b5b932f9a344d4","934c0d476e0913dedfc11ab20974d89e"],"score":25.47}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548}],"score":34.12,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":450},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":450},"family":"MMMU-Pro","row_ids":["681af2b5c57d4ca46641c38951c3d479"],"score":26.47}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069}],"score":26.47,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":49.15,"percentile_exact":{"denominator":59,"numerator":2900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","peer_count":64,"percentile":26.98,"percentile_exact":{"denominator":63,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","peer_count":52,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","peer_count":82,"percentile":24.69,"percentile_exact":{"denominator":81,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":78,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","peer_count":24,"percentile":17.39,"percentile_exact":{"denominator":23,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","peer_count":50,"percentile":18.37,"percentile_exact":{"denominator":49,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069}],"result_updated_at":null,"score":27.6,"score_exact":{"denominator":19177971701616,"numerator":529374170552879}}],"release_date":"2026-08-10","release_status":"published","release_verified":true,"source_id":"meta/muse-glimmer-30b"},"m-c0580b01221ed260":{"id":"m-c0580b01221ed260","in_window":true,"name":"Muse Spark 1.3","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"75958d551dae8b1712ddad1f31e0ada8","model_id":"m-c0580b01221ed260","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5868686868686869,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3a808fea0ebda449aa1bb91e"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"3a808fea0ebda449aa1bb91e","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5868686868686869},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","model_id":"m-c0580b01221ed260","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","model_id":"m-c0580b01221ed260","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","model_id":"m-c0580b01221ed260","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"official_evidence":{"artifact_sha256":"b221b103503442df9d920dbb3ff2002c3c9c86755300e3f9635018eaea8c2e56","basis":"official_release_page","date":"2026-09-02","sha256":"b221b103503442df9d920dbb3ff2002c3c9c86755300e3f9635018eaea8c2e56","url":"https://research.meta.ai/blog/introducing-muse-spark-1-3"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":41.67,"evaluator_count":2,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["6e4fcbf667751f178fabfb4106350208"],"score":31.58},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":25},"family":"HLE","row_ids":["5a5d1e52735d8efc5b98d11b601491a8"],"score":25.0}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"Diamond","peer_count":17,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4}],"score":28.29,"weight":25},{"effective_weight_percent":41.67,"evaluator_count":2,"exact":{"denominator":16,"numerator":975},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":175},"family":"Terminal-Bench","row_ids":["73167c475c92b2f6b68a011a68a996ab"],"score":21.88},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"AA-Briefcase","row_ids":["0619cb6c1df6def76a850c294e50de63"],"score":100.0}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"v1.1","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162}],"score":60.94,"weight":25},{"effective_weight_percent":16.67,"evaluator_count":1,"exact":{"denominator":27,"numerator":1900},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1900},"family":"Humanity Sixth Sense","row_ids":["fcb2b9f9c2c7859d3369fef86db555f4"],"score":70.37}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","peer_count":28,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4}],"score":70.37,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"Diamond","peer_count":17,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"v1.1","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","peer_count":28,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4}],"result_updated_at":null,"score":48.91,"score_exact":{"denominator":98496,"numerator":4817075}}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"source_id":"meta/muse-spark-1.3"},"m-c0755395b2da9721":{"id":"m-c0755395b2da9721","in_window":true,"name":"Nex-N2-Pro","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","model_id":"m-c0755395b2da9721","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","model_id":"m-c0755395b2da9721","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":4200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4200},"family":"GPQA Diamond","row_ids":["004e3b6a13ca6828e88e6f4e76a97e76"],"score":71.19}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":71.19,"percentile_exact":{"denominator":59,"numerator":4200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192}],"score":71.19,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3081078,"numerator":117236125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["0456438c8b02eed2e908e4593e41dcf0"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3700},"family":"GDPval-AA","row_ids":["9050cb18d0244c05aa4a3bdecef069eb"],"score":45.68},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":700},"family":"AutomationBench","row_ids":["c9e2b1db45e7b21cbd9a431702d99eb1"],"score":36.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2700},"family":"AA-Briefcase","row_ids":["82a245d0d5111b9661f660b2a94b3db6"],"score":35.06}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v2.1","peer_count":82,"percentile":45.68,"percentile_exact":{"denominator":81,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v1.1","peer_count":78,"percentile":35.06,"percentile_exact":{"denominator":77,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26}],"score":38.05,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":125},"family":"IFBench","row_ids":["353142d52584e4a6aac9d52541e06d23"],"score":62.5},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2225},"family":"GDP.pdf","row_ids":["312a7fa68250f3fcf7a32404595e6dc4"],"score":58.55}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126}],"score":60.53,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":71.19,"percentile_exact":{"denominator":59,"numerator":4200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v2.1","peer_count":82,"percentile":45.68,"percentile_exact":{"denominator":81,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v1.1","peer_count":78,"percentile":35.06,"percentile_exact":{"denominator":77,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126}],"result_updated_at":null,"score":56.31,"score_exact":{"denominator":2544970428,"numerator":143298061675}}],"release_date":"2026-06-02","release_status":"published","release_verified":true,"source_id":"nex-agi/nex-n2-pro"},"m-c46d8f7f1c3144cb":{"id":"m-c46d8f7f1c3144cb","in_window":true,"name":"K-EXAONE 2.0 0803","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":83020675},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2650},"family":"GPQA Diamond","row_ids":["c948234960090e8c9ef801d4c2e468fc"],"score":44.92},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["5d3cac863b9cb777dcf0429fca5619c7"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1150},"family":"HLE","row_ids":["1e3489d548cb804c3cfe4ff52df120e8"],"score":18.25},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":350},"family":"CritPt","row_ids":["55904790fd4c7a2e974218f474b3d773"],"score":20.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":44.92,"percentile_exact":{"denominator":59,"numerator":2650},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04}],"score":25.27,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":73359,"numerator":1386274},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":225},"family":"tau3-Banking","row_ids":["77cc539d6e52970112947681c17a7cf5"],"score":17.31},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":700},"family":"GDPval-AA","row_ids":["3eb5f8707daab0412d08dc253ada641e"],"score":25.93},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["017b50b4eb6f553388684608c12adbeb"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":375},"family":"AutomationBench","row_ids":["cf48341c74514d20a48b2f428df05755"],"score":19.74},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":200},"family":"AA-Briefcase","row_ids":["42d65fa891e8c3a4f35b4ea031798999"],"score":18.18}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","peer_count":82,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":78,"percentile":18.18,"percentile_exact":{"denominator":11,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09}],"score":18.9,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":500},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":500},"family":"FrontierCode","row_ids":["a86b3f52322966e22c9f2052d8fbc893"],"score":10.2}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","peer_count":50,"percentile":10.2,"percentile_exact":{"denominator":49,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404}],"score":10.2,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":119725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":825},"family":"GDP.pdf","row_ids":["7e1c74ae3bf26c2c5b56bd7d43f2a5df"],"score":21.71},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2000},"family":"AA-Omniscience","row_ids":["6b5784505206d57d0bba2ceb5b8de557","3bcf44016a4447d1427e1bf692c1e1bc"],"score":37.74}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963}],"score":29.72,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":44.92,"percentile_exact":{"denominator":59,"numerator":2650},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","peer_count":82,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":78,"percentile":18.18,"percentile_exact":{"denominator":11,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","peer_count":50,"percentile":10.2,"percentile_exact":{"denominator":49,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963}],"result_updated_at":null,"score":21.14,"score_exact":{"denominator":13758110133768,"numerator":290851131141355}}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"source_id":"published/aa:k-exaone-2-0-0803"},"m-c58d5bdb4d2e38bc":{"id":"m-c58d5bdb4d2e38bc","in_window":true,"name":"K2 Horizon 7B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":39117,"numerator":787925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1850},"family":"GPQA Diamond","row_ids":["885831acf33b9d6286fb559cc7b4ec0e"],"score":31.36},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d166f2e488fa71112bb60e275a7535f1"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["3441fc8b1a39080bc19e5b920fef2c87"],"score":11.76}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":20.14,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1100385,"numerator":49356484},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":800},"family":"tau3-Banking","row_ids":["33c50c79dd62f0151f2db3df92a057bc"],"score":61.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2900},"family":"GDPval-AA","row_ids":["4aa9f9d8e559b9430ab1f871a4ca4271"],"score":35.8},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["5a757d2c757812d80a306fe7007417cd"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":750},"family":"AutomationBench","row_ids":["b87ec7052a30aebc07b1b22c9deef3cd"],"score":39.47},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":500},"family":"AA-Briefcase","row_ids":["6d3a1ea10b80de3baf1922e2ef5fc22e"],"score":45.45}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","peer_count":82,"percentile":35.8,"percentile_exact":{"denominator":81,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","peer_count":78,"percentile":45.45,"percentile_exact":{"denominator":11,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51}],"score":44.85,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":800},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":800},"family":"FrontierCode","row_ids":["d64bab068e454f76d91e1fb18d73dd68"],"score":16.33}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","peer_count":50,"percentile":16.33,"percentile_exact":{"denominator":49,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525}],"score":16.33,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":625},"family":"GDP.pdf","row_ids":["48f1c6876fabfb820ee570e743c9611a"],"score":16.45}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"score":16.45,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","peer_count":82,"percentile":35.8,"percentile_exact":{"denominator":81,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","peer_count":78,"percentile":45.45,"percentile_exact":{"denominator":11,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","peer_count":50,"percentile":16.33,"percentile_exact":{"denominator":49,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"result_updated_at":null,"score":25.34,"score_exact":{"denominator":194690237742,"numerator":4933003274473}}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"source_id":"published/aa:k2-horizon-7b"},"m-c65e3e17492500e9":{"id":"m-c65e3e17492500e9","in_window":true,"name":"GLM-5.3","observed_records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GLM-5.3","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"9a912208deecc3da383c31d8417739d7","model_id":"m-c65e3e17492500e9","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":37.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column4","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.9},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"bbb96b1ec88d510cf9f1da8d736fb7f4","model_id":"m-c65e3e17492500e9","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":66.9,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"66.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column4","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":66.9},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GLM-5.3","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"28ec00bcc7a9b0d7e2be777a2deb9b0e","model_id":"m-c65e3e17492500e9","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":48.8,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"48.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column4","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":48.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GLM-5.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"4f28c1e35ee6e421b89e2a85fbc8ba80","model_id":"m-c65e3e17492500e9","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column GLM-5.3. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":88.1,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"88.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column4","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":88.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"7bacdcf72f2524d7ecbbadf7b0f6e174","model_id":"m-c65e3e17492500e9","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":91.4,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"91.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":91.4},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"c0d551d2a17a8d88a7eef965f627a78f","model_id":"m-c65e3e17492500e9","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":38.1,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"38.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"083544e7fa726861ec84d82ed3ea9450","model_id":"m-c65e3e17492500e9","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":49.4,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"49.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":49.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","model_id":"m-c65e3e17492500e9","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"glm-5.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"f0c4c8c8946a2cc602d15226e2c11a49","model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":16.4,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f42675d05d3dc080cf8118d0"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f42675d05d3dc080cf8118d0","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":16.4},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","model_id":"m-c65e3e17492500e9","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","model_id":"m-c65e3e17492500e9","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0d42916648a5d251cec213581173bb2b","model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1185.12,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1185.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"83f0cd9fed740460c07252b458b9b686","model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.319,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"436f19d006eb9d5593af32cf7feebeaf","model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3675966715614293,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3675966715614293},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba5c747e026e0c62f82c17a38fa5481c","model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.519790983751346,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.519790983751346},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"fecc9e596509bb91af4290ca53041611","model_id":"m-c65e3e17492500e9","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0380952380952381,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0380952380952381},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c5bdba52f2dfb5d75aa9320af9dd9f9","model_id":"m-c65e3e17492500e9","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0952380952380952,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0952380952380952},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a828660ab96f0cc0b2bdba0a5e5a48c0","model_id":"m-c65e3e17492500e9","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.127379797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.127379797979798},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","model_id":"m-c65e3e17492500e9","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"992e41c99a19582370117173efb9c4eb","model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.175,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"54ce2558ad375331e00840130efd4dfb","model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.088,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6da70947470fbcfc9486b425d16e0547","model_id":"m-c65e3e17492500e9","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.348484848484849,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"glm-5-3-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348484848484849},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa12d2968deb06d22cebe6ac9936af8","model_id":"m-c65e3e17492500e9","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.419191919191919,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"glm-5-3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.419191919191919},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4933726aa62552e1e851d094c416d65e","model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.215,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.215},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"999080053b7866e42174b5010a56f2de","model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1333.34,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1333.34},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155}],"official_evidence":{"artifact_sha256":"2dd162815ce7015e44d62c786bf3b76ffde8b7dc7834b1cd6a791f78c3d68667","basis":"official_release_page","date":"2026-08-14","sha256":"2dd162815ce7015e44d62c786bf3b76ffde8b7dc7834b1cd6a791f78c3d68667","url":"https://www.zhipuai.cn/zh/research/162"},"primary_profile_id":"native","profiles":[{"benchmark_count":19,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":1200591,"numerator":76127425},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5100},"family":"GPQA Diamond","row_ids":["85f2af62bad122fb48d79f2a5cd0ba76"],"score":86.44},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["6cbde617319ab4da7917bf56251c3a2e"],"score":44.74},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3700},"family":"HLE","row_ids":["20084307dcc6cac0bc9c76758b5e2139"],"score":58.73},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3250},"family":"CritPt","row_ids":["6177f516623a8dfc4d1c35316f9c20f2"],"score":63.73}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":86.44,"percentile_exact":{"denominator":59,"numerator":5100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","peer_count":64,"percentile":58.73,"percentile_exact":{"denominator":63,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335}],"score":63.41,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":4,"exact":{"denominator":196240968,"numerator":14270932625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":250},"family":"APEX-Agents","row_ids":["21a8aa59fcb39de6e961594401b8acf8"],"score":83.33},{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":100},"family":"tau3-Banking","row_ids":["0034bad285685afff6ae0af098ac3976"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7000},"family":"GDPval-AA","row_ids":["8faa2a90e5a7df630665e90e7dc8da8c"],"score":86.42},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1675},"family":"Terminal-Bench","row_ids":["379465a1dce0780ec05eea7fd360c800"],"score":69.79},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1525},"family":"AutomationBench","row_ids":["a4b22debd5eda444528d12c717ed41a1"],"score":80.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6000},"family":"AA-Briefcase","row_ids":["98953845b6332ba7948a61f814a750d3"],"score":77.92},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["4a39f12346421fa0779ea6e76813ff8f"],"score":33.33},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":700},"family":"MCP Atlas","row_ids":["442d36ffd8ae4e2ebc285d09f51efdf2"],"score":77.78},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1050},"family":"AA-AnalystAgent","row_ids":["18b8a0177c3660f9aaf285f2b3571603"],"score":45.65}],"family_count":9,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","peer_count":82,"percentile":86.42,"percentile_exact":{"denominator":81,"numerator":7000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":78,"percentile":77.92,"percentile_exact":{"denominator":77,"numerator":6000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"score":72.72,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":273,"numerator":16000},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":400},"family":"FrontierCode","row_ids":["326d68113ecd602bd7d85d3d2b454d3f"],"score":57.14},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":800},"family":"Vals Code Migration","row_ids":["78dbe912ae9dbb4dfacad05113fef49e"],"score":61.54},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":400},"family":"DeepSWE","row_ids":["a1f523e170da53cbcf5ad3cbcb817e6a"],"score":57.14}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"2.0","peer_count":50,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1.1","peer_count":27,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","included":true,"model_id":"m-c65e3e17492500e9","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1}],"score":58.61,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":34238,"numerator":1354975},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["ba325e0068d64ba19ce176fe95dbb3e4"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":400},"family":"SimpleQA Verified","row_ids":["dd5c55d1f80c2ba62c676b7e3e150a64"],"score":23.53},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2500},"family":"AA-Omniscience","row_ids":["be51199fdc7d1731045d427201852660","330a13ca3aada5b9674ba75668e0acfc"],"score":47.17}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838}],"score":39.58,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":86.44,"percentile_exact":{"denominator":59,"numerator":5100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","peer_count":64,"percentile":58.73,"percentile_exact":{"denominator":63,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","peer_count":82,"percentile":86.42,"percentile_exact":{"denominator":81,"numerator":7000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":78,"percentile":77.92,"percentile_exact":{"denominator":77,"numerator":6000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"2.0","peer_count":50,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1.1","peer_count":27,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","included":true,"model_id":"m-c65e3e17492500e9","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838}],"result_updated_at":"2026-08-28","score":59.63,"score_exact":{"denominator":2441081826591408,"numerator":145567376505820175}}],"release_date":"2026-08-14","release_status":"published","release_verified":true,"source_id":"zhipuai/glm-5.3"},"m-c947385433db3f03":{"id":"m-c947385433db3f03","in_window":true,"name":"Qwen3.8 27B","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","model_id":"m-c947385433db3f03","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","model_id":"m-c947385433db3f03","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 27B / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"d437bf4a69b6a24c59bada231457a5bc","model_id":"m-c947385433db3f03","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"0a12251bcdc69252d8e0391062c04efe","model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.699421965317919,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.699421965317919},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7c0aa7b7706bacb6c14407a8a7a9f461","model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.741618497109827,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.741618497109827},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"5ca8b7088f0eafb66b08f6082fa16296","model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.73757225433526,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.73757225433526},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e963a991d26d29a030553cf253aa8499","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":900.84,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b-non-reasoning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":900.84},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc4ab5cd5a57325a1ea530c7d42ceb8b","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1371.79,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1371.79},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24cdeca0588edc5fc8a2271fa603fe1a","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1269.55,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1269.55},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6bbb04cccd1e43c083784b1650fd275a","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.091,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b-non-reasoning:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"21c921f8950b8d0002ea46ae123ee958","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.17866666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17866666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9f4f11f914d3ab5295651ee50040f4df","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.183,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.183},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5844b72745de9798297f3b9e7a24b3ac","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.81994866153282,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b-non-reasoning:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.81994866153282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e298eebd839f0666980d289ba85f5653","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.34577922077922074,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34577922077922074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f1accb8dbba25344b74774765b16ec0b","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5132598939208486,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5132598939208486},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","model_id":"m-c947385433db3f03","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8737dc80c63db6e6987d64ba777be822","model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.12434667722683214,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b-non-reasoning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12434667722683214},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d97d4395a989a9285a92552b07484a1e","model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.40252101015938635,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40252101015938635},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8307c7633c9f974d8ce2526d440bfdd7","model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3010329530343377,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3010329530343377},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c20a9dd8cd5adb1c62ba7310fa9ddb0","model_id":"m-c947385433db3f03","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b-non-reasoning:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"017fe7274db4b5e5e777a432b4886f3f","model_id":"m-c947385433db3f03","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea9242a24574cd6edb74236baf2f2a9","model_id":"m-c947385433db3f03","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","model_id":"m-c947385433db3f03","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ba0100e2f0dfb2254ad6610afb255d89","model_id":"m-c947385433db3f03","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0371893939393939,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0371893939393939},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2a58e9d0add012cd5f60c96e79d87d38","model_id":"m-c947385433db3f03","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.115037626262626,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115037626262626},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"5a7079def18d458c9373e6d133019e26","model_id":"m-c947385433db3f03","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.123696212121212,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.123696212121212},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","model_id":"m-c947385433db3f03","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2696fb96fac279e820374f5935f055de","model_id":"m-c947385433db3f03","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b-non-reasoning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5a9ac5c057bf7f22ddb3c0f0c004a86b","model_id":"m-c947385433db3f03","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.067,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.067},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b9d26d34bedd63dd22d99c28886a9508","model_id":"m-c947385433db3f03","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.068,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.068},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","model_id":"m-c947385433db3f03","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8395e43be344be658905775f649b084d","model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b-non-reasoning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9e0f871dc000a146ae77cfcc0caf9b6","model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"217c868a4b0108feba5aa58a275f6fe0","model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.104,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.104},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"060b4f0030cb2f4d8abf7fde91641a2f","model_id":"m-c947385433db3f03","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b-non-reasoning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7063a6b5f336f5eff87f370e1e93b388","model_id":"m-c947385433db3f03","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0505050505050505},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c288cc302ca0055a675eed3d1700dc0e","model_id":"m-c947385433db3f03","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0252525252525253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0252525252525253},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","model_id":"m-c947385433db3f03","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19b0e2cbfac5dda255748dc035aa572a","model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.02,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b-non-reasoning:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f02aa2a50ec5c65112bb8266920c78d0","model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.165,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.165},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02d2311a9c42d7b6d866d9e045a6582d","model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.055,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.055},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"862e30c125143b50335c08b196c3563e","model_id":"m-c947385433db3f03","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1082.06,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b-non-reasoning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1082.06},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe28efd15ca4f984963b0bfffbecf250","model_id":"m-c947385433db3f03","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1388.62,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1388.62},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5264c213b70ffa8280024c357f90b78d","model_id":"m-c947385433db3f03","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1370.26,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1370.26},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","model_id":"m-c947385433db3f03","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","model_id":"m-c947385433db3f03","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"59641e983c8baf60f3afa400154320d7","model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.818181818181818,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.818181818181818},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e1db2a4a79e56a32fdcdd90f8539e5d1","model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b-medium:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0e776f0b96d97c744ba8771be1d8d49c","model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b-low:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"79bc83d205a370cf9d4ed71ef0ca0b4b","model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.2,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-non-reasoning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b-non-reasoning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5aa79f4b40af631a1da57f82cc2ccb41","model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.474226804123711,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-medium:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b-medium:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.474226804123711},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"625405886814a9eb9d6b86f64ac262b9","model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.321649484536082,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.321649484536082},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-14","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":6571656,"numerator":314311525},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4750},"family":"GPQA Diamond","row_ids":["ba847bf9e90b2e6c2d3a1b021f6f1ff2"],"score":80.51},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["2f71320c4fe6daaec73fc9a09a2e75a6"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3200},"family":"HLE","row_ids":["ee3f7405021072ff698c63b6e1fa510f"],"score":50.79},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":350},"family":"CritPt","row_ids":["a3193e796745f7cc574aae69daceb5f1"],"score":20.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","included":true,"model_id":"m-c947385433db3f03","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond","peer_count":64,"percentile":50.79,"percentile_exact":{"denominator":63,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04}],"score":47.83,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":220077,"numerator":14695565},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1250},"family":"tau3-Banking","row_ids":["02b14b2382923c4835faa17152c2d94b"],"score":96.15},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1700},"family":"GDPval-AA","row_ids":["022f63e78e1f4658e8af4b434e595fa3"],"score":62.96},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":160},"family":"Terminal-Bench","row_ids":["bb15f242cbb67b0a52ecb1efaeaf802e"],"score":53.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1100},"family":"AutomationBench","row_ids":["07fee1c32e6f2287a67ddf683ea02c91"],"score":57.89},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":700},"family":"AA-Briefcase","row_ids":["bee797f9fafa487f710b7abf6c3fddcf"],"score":63.64},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"EnterpriseOps-Gym","row_ids":["cea12af74b5609df87befba4292e8ab2"],"score":66.67}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","included":true,"model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2.1","peer_count":82,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","included":true,"model_id":"m-c947385433db3f03","original_version_label":"4.0","peer_count":76,"percentile":53.33,"percentile_exact":{"denominator":3,"numerator":160},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","included":true,"model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":78,"percentile":63.64,"percentile_exact":{"denominator":11,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA run","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471}],"score":66.77,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":21,"numerator":800},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":300},"family":"FrontierCode","row_ids":["328d104cb62d50eef3ae2e92cc4db522"],"score":42.86},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":100},"family":"DeepSWE","row_ids":["b993c11069821aca78f26acbce738dbc"],"score":33.33}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","included":true,"model_id":"m-c947385433db3f03","original_version_label":"2.0","peer_count":50,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","included":true,"model_id":"m-c947385433db3f03","original_version_label":"1.1","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2}],"score":38.1,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":9063,"numerator":536350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"IFBench","row_ids":["860832a18411b625ebef7842433113ca"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1300},"family":"GDP.pdf","row_ids":["4189c00054cffddacad0958de8d0f156"],"score":68.42},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["5ce6a4a59546a3ac7bd2230a498f7f48","ee80bfee31969adf86abcbeaae10d4e4"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current published protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742}],"score":59.18,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":600},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":600},"family":"MMMU-Pro","row_ids":["fa359632aae944ad7f900cc61aa55351"],"score":35.29}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821}],"score":35.29,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","included":true,"model_id":"m-c947385433db3f03","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond","peer_count":64,"percentile":50.79,"percentile_exact":{"denominator":63,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","included":true,"model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2.1","peer_count":82,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","included":true,"model_id":"m-c947385433db3f03","original_version_label":"4.0","peer_count":76,"percentile":53.33,"percentile_exact":{"denominator":3,"numerator":160},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","included":true,"model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":78,"percentile":63.64,"percentile_exact":{"denominator":11,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA run","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","included":true,"model_id":"m-c947385433db3f03","original_version_label":"2.0","peer_count":50,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","included":true,"model_id":"m-c947385433db3f03","original_version_label":"1.1","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current published protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821}],"result_updated_at":"2026-08-26","score":51.64,"score_exact":{"denominator":2620592406432,"numerator":135314940818345}}],"release_date":"2026-08-14","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.8-27b"},"m-caf1ef1f3bdf5452":{"id":"m-caf1ef1f3bdf5452","in_window":true,"name":"DeepSeek V4.1 Flash 1017","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"65d18ec825f22a5bf73b53126498f464","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.691907514450867,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691907514450867},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"975dc845b85346f78b00a766be31b8ea","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797109826589595,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797109826589595},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a9c9f93aa794c9a3fcd167406210fd0","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.32,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.32},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0ddc521d18c763b1e011aa82f42d16","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1415.7,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1415.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d389d661688b3a9f75a1d3dc17fc60a1","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21866666666666668,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21866666666666668},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2a25754143f9024b0018cbc2c9f1e49b","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3893333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3893333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"620a073a8d8620b211d1026dd05f157f","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7470136518771331,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7470136518771331},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1f5a3d7104306555b6391d832ea1aca5","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38755458515283847,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38755458515283847},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efeaf909a866f14a9089212f7876fb00","model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4877893544529126,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4877893544529126},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"26f38dbf96eff4351a310f622e79da7b","model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849296525816153,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849296525816153},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8841ce02079360f927638c98132c0b33","model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0238095238095238,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0238095238095238},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2736951c2ff4d38eb65bfdd6526592d","model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.104761904761905,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.104761904761905},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2bdec8d61778c54030041dfd425aa3b","model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.060984595959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.060984595959596},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"763c12252b9548502d33d4f214d5052b","model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.168437878787879,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.168437878787879},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cbd0e9e35b3ef69b6b8ee1bd713a64fa","model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.071,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.071},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"734956c3228c7619bb1dca47baa5c0cb","model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.208,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"20b7c87d5464baec8ebe5c091cfefd68","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.044,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.044},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"391d19d60651185403b08ca595fa4899","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.142,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e2c92f63a196dfe970268dbe4753317","model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0909090909090909,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0909090909090909},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"343f2f2532cc9f61cc636915abf75a69","model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.287878787878788,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287878787878788},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e62fa5ce7d9f527df294b117e42de68a","model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.338383838383838,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-1-flash-1017:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.338383838383838},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96c7ea64af8d914523dd4f7c5117b5c8","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7bc4d787eb3400ad5d8e493fa4e84a31","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.275,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-17","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":37128,"numerator":2162375},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":350},"family":"FrontierMath","row_ids":["f2e8485b9221333706b0e9e14cd8c4e7"],"score":38.89},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1975},"family":"Terminal-Bench-Science","row_ids":["ebff4c97ce1c3d30b95d6d855419e5c1"],"score":75.96},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4600},"family":"HLE","row_ids":["2bfa42798a174eb5a88b71f065cd4aef"],"score":73.02},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2300},"family":"CritPt","row_ids":["d9b68c289d550eec511fa758a53ee05a"],"score":45.1}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","peer_count":64,"percentile":73.02,"percentile_exact":{"denominator":63,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","peer_count":52,"percentile":45.1,"percentile_exact":{"denominator":51,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25}],"score":58.24,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114912,"numerator":9263075},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2500},"family":"GDPval-AA","row_ids":["0a034403545d9a6f875f18bdfdb8614b"],"score":92.59},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1375},"family":"Terminal-Bench","row_ids":["6d06cabd5082651a9ea6ff6f7709c136"],"score":57.29},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1650},"family":"AutomationBench","row_ids":["74476994a8c12c8ddef6748c439b2908"],"score":86.84},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":600},"family":"AA-Briefcase","row_ids":["e9014188fe35aeb589dfba302eb5ab58"],"score":85.71}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","peer_count":82,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":78,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42}],"score":80.61,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":49200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3800},"family":"FrontierCode","row_ids":["981d7322c13b53f8d5edfa1fdb6e95e5"],"score":77.55},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1000},"family":"Vals Code Migration","row_ids":["1080a38736b6ba3768e68ace6a340f30"],"score":76.92}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","peer_count":50,"percentile":77.55,"percentile_exact":{"denominator":49,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","peer_count":27,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52}],"score":77.24,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2456375},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1325},"family":"GDP.pdf","row_ids":["b06f9a392143fecd0d26f4f434ebd403"],"score":69.74},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":500},"family":"SimpleQA Verified","row_ids":["8e35d6850446c900d7c667bcebf2d5eb"],"score":29.41},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2350},"family":"AA-Omniscience","row_ids":["28843797b0f403bc00b3b129a0250f86","cd12ad2957c69fc5e8660a959eaae5db"],"score":44.34}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795}],"score":47.83,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":153,"numerator":8575},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"Humanity Sixth Sense","row_ids":["60e8851ddeb919672b146036d9d51182"],"score":44.44},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1150},"family":"MMMU-Pro","row_ids":["33026de8036ee6188f530a53e9aaa0aa"],"score":67.65}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","peer_count":28,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289}],"score":56.05,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","peer_count":64,"percentile":73.02,"percentile_exact":{"denominator":63,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","peer_count":52,"percentile":45.1,"percentile_exact":{"denominator":51,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","peer_count":82,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":78,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","peer_count":50,"percentile":77.55,"percentile_exact":{"denominator":49,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","peer_count":27,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","peer_count":28,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289}],"result_updated_at":"2026-10-19","score":65.33,"score_exact":{"denominator":37686999168,"numerator":2462117234705}}],"release_date":"2026-10-17","release_status":"published","release_verified":true,"source_id":"deepseek/deepseek-v4.1-flash-1017"},"m-cccbc1ce3042d216":{"id":"m-cccbc1ce3042d216","in_window":true,"name":"Claude Opus 4.8","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"8bd910e44b3e0be43d3313d301ed4e94","model_id":"m-cccbc1ce3042d216","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":80.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":80.0},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","model_id":"m-cccbc1ce3042d216","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Opus4.8","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"5b6ba99bf6e1deff4ca257f908136e39","model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":59.0,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"59.0","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column1"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column1","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":59.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Opus4.8","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d58fa04ae1f8f5897c36ae81b8ac3dcf","model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":27.2,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"27.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column1"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column1","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":27.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Opus4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"6d5a13969cd11982fa7a1306707181d9","model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":62.2,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"62.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column1"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column1","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":62.2},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Opus4.8","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d48870e5e71cd2cc3b689552b141ee4a","model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":75.6,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"75.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column1"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column1","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":75.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Opus4.8","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"c316ffaf81b4d0057dc56f75f0232808","model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Opus4.8. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.0,"research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.0","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column1"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column1","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"ec4b61209e562c5ed86fc0f393950588","model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":59.0,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"59.0","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":59.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"d58257a11e2e15bf0bb878bb097f91ba","model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.2,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"c0d394b3326f2e9c0afb7d54483e968c","model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.0,"research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.0","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.0},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"04a252c77bb5bb362fedd335d947b332","model_id":"m-cccbc1ce3042d216","original_version_label":"1.1; explicit developer footnote","profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":59.0,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"59.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":59.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"682b2a95186a2a4bc27f60c8b3158f30","model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":27.2,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"27.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":27.2},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b730b9240ece9fc9ae2c909c6d9b5b27","model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.0,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"2ba9fb0fcdacf536fe9c61d211eab736","model_id":"m-cccbc1ce3042d216","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":83.6,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"83.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":83.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"official_evidence":{"artifact_sha256":"f43b616507c976c560385970d34c456a9634c2e81776613c8670c64ab10e5e35","basis":"official_release_page","date":"2026-05-28","sha256":"f43b616507c976c560385970d34c456a9634c2e81776613c8670c64ab10e5e35","url":"https://www.anthropic.com/news/claude-opus-4-8"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":19,"numerator":200},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":0},"family":"FrontierMath","row_ids":["040f48951e0ebb9ff6bfbde83c816f48"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":0},"family":"GPQA Diamond","row_ids":["29bbe2be4ccedeb019b3ccf6a3404eb9"],"score":0.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["540adb8175ab28d08c531a963e348646"],"score":31.58}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v2","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":10.53,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":5,"exact":{"denominator":3,"numerator":124},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"APEX-Agents","row_ids":["9584a53c71e055a6a275ffd8975b0b25"],"score":40.0},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":275},"family":"Terminal-Bench","row_ids":["2539da8c839981606943a4efe846b325"],"score":45.83},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"AutomationBench","row_ids":["6bd66300b839f4ff34d5fd8249350449"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"AA-Briefcase","row_ids":["52ae106aa8230eabcaac4637b6d547ea"],"score":50.0},{"benchmark_version":"1000 tasks","exact":{"denominator":3,"numerator":100},"family":"MCP Atlas","row_ids":["4cec2c81b7ca1a60e4ad36412d5896a7"],"score":33.33}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","peer_count":49,"percentile":45.83,"percentile_exact":{"denominator":6,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2}],"score":41.33,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["dee93fd91e74e8ba6728dba7ca4fd982"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":459,"numerator":11950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"IFBench","row_ids":["0a5ea9e2a9d5f26a6a034bb3853216a5"],"score":22.22},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"GDP.pdf","row_ids":["dc77f0ff3c077230aef1b2b8cb04ec92"],"score":0.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":950},"family":"SimpleQA Verified","row_ids":["dbd9d582709dbd41201aa5f60842dfb0"],"score":55.88}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53}],"score":26.03,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":9,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["2c5710356bbeeb960760bc793dbd6d80"],"score":11.11},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["24b2d438f8c9ef458157a109e7bcb701"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","peer_count":28,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9}],"score":5.56,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v2","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","peer_count":49,"percentile":45.83,"percentile_exact":{"denominator":6,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","peer_count":28,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9}],"result_updated_at":"2026-08-27","score":18.73,"score_exact":{"denominator":8721,"numerator":163322}}],"release_date":"2026-05-28","release_status":"published","release_verified":true,"source_id":"anthropic/claude-opus-4-8"},"m-cf177eb625d4bd3e":{"id":"m-cf177eb625d4bd3e","in_window":true,"name":"GLM-5.3-Flash","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4e220e5978b3977fd57d2c06a76a9766","model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eca9bd0f6ebf44ed64158c0a214f9271","model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"glm-5-3-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1642914,"numerator":100299025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4900},"family":"GPQA Diamond","row_ids":["3551c30b97ba4cb5f0fc94e087d94b59"],"score":83.05},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":750},"family":"Terminal-Bench-Science","row_ids":["0ca35771477057319be8d71ddeaad930"],"score":57.69},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3800},"family":"HLE","row_ids":["a57cb2704c93293d339e91362f40ae46"],"score":60.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2200},"family":"CritPt","row_ids":["620b6f3e6314727d16068f5503869955"],"score":43.14}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":83.05,"percentile_exact":{"denominator":59,"numerator":4900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","peer_count":64,"percentile":60.32,"percentile_exact":{"denominator":63,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","peer_count":52,"percentile":43.14,"percentile_exact":{"denominator":51,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245}],"score":61.05,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":141729588,"numerator":8766837125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":2375},"family":"tau3-Banking","row_ids":["9a44264bd8b89c341b6a8bd9496be3bf"],"score":91.35},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6800},"family":"GDPval-AA","row_ids":["3091d538e8d1e5edf6eea60314cea6a6"],"score":83.95},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":775},"family":"Terminal-Bench","row_ids":["28abab0f2a19e622cf15ffcdc9be3dfd"],"score":64.58},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1475},"family":"AutomationBench","row_ids":["e492bf3e10afd1293a1152cd005f96a3"],"score":77.63},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":800},"family":"AA-Briefcase","row_ids":["3015c7ad3cd2d55a621f41dcebd98f3d"],"score":72.73},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["b8b704019523524af0ce859e89927fa0"],"score":16.67},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":600},"family":"AA-AnalystAgent","row_ids":["a9057fe2d742aeb1aa0666773aa5958d"],"score":26.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","peer_count":82,"percentile":83.95,"percentile_exact":{"denominator":81,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","peer_count":49,"percentile":64.58,"percentile_exact":{"denominator":12,"numerator":775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":77.63,"percentile_exact":{"denominator":19,"numerator":1475},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":78,"percentile":72.73,"percentile_exact":{"denominator":11,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","peer_count":24,"percentile":26.09,"percentile_exact":{"denominator":23,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25}],"score":61.86,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":30950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2500},"family":"FrontierCode","row_ids":["32536ff0b3b8576673f9f05323026239"],"score":51.02},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":600},"family":"Vals Code Migration","row_ids":["efd99741d3524e2d5ab3b679f8da8006"],"score":46.15}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","peer_count":50,"percentile":51.02,"percentile_exact":{"denominator":49,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","peer_count":27,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65}],"score":48.59,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":108625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1225},"family":"GDP.pdf","row_ids":["f95dbe30b012b632cab81f4598931983"],"score":64.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["a76a9fe9ce50e17717765fcaebd6fbf8","7e86ee60903a41ed29953982d04673a1"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499}],"score":53.93,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["7d2a87235efadfc965d04473b687dd29"],"score":33.33}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","peer_count":28,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8}],"score":33.33,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":83.05,"percentile_exact":{"denominator":59,"numerator":4900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","peer_count":64,"percentile":60.32,"percentile_exact":{"denominator":63,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","peer_count":52,"percentile":43.14,"percentile_exact":{"denominator":51,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","peer_count":82,"percentile":83.95,"percentile_exact":{"denominator":81,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","peer_count":49,"percentile":64.58,"percentile_exact":{"denominator":12,"numerator":775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":77.63,"percentile_exact":{"denominator":19,"numerator":1475},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":78,"percentile":72.73,"percentile_exact":{"denominator":11,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","peer_count":24,"percentile":26.09,"percentile_exact":{"denominator":23,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","peer_count":50,"percentile":51.02,"percentile_exact":{"denominator":49,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","peer_count":27,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","peer_count":28,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8}],"result_updated_at":null,"score":54.56,"score_exact":{"denominator":210957688717776,"numerator":11510724316142215}}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"source_id":"zhipuai/glm-5.3-flash"},"m-cfe0652e9d685783":{"id":"m-cfe0652e9d685783","in_window":true,"name":"Claude Opus 5.5","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench 4.0 (AA)","benchmark_version":"4.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"bbe3b0185e63421d7e8bf29fa803e68e","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench 4.0 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"6dfc6fa538421f2fee6d35c4","raw_value":59.6,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"59.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench 4.0 (AA):Claude Opus 5.5"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench:Terminal-Bench 4.0 (AA)","source_record_id":"carbon-image:Terminal-Bench 4.0 (AA):Claude Opus 5.5","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":59.6},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"c81051296c04b2de65d68a3054256560","model_id":"m-cfe0652e9d685783","original_version_label":"2.1","profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":87.2,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"87.2%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Claude Opus 5.5"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Claude Opus 5.5","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":87.2},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench Zapier","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"c4b221a2cb6a2006aac48962128f37ff","model_id":"m-cfe0652e9d685783","original_version_label":"current published developer protocol","profile_key":"native","protocol_description":"Official comparison image; AutomationBench Zapier; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"f658b6663efcee3dc42343bd","raw_value":42.5,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"42.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:AutomationBench Zapier:Claude Opus 5.5"},"source_name":"blog.google","source_protocol_key":"official:blog.google:AutomationBench:AutomationBench Zapier","source_record_id":"carbon-image:AutomationBench Zapier:Claude Opus 5.5","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":42.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"6d4e060e399a72c43217ec5b1d6a09b6","model_id":"m-cfe0652e9d685783","original_version_label":"current published dataset","profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":66.3,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"66.3%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Claude Opus 5.5"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Claude Opus 5.5","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":66.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science 0.1 (AA)","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","dimension":"reasoning","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"47678bec4f89a1e5999bbadc4fe96b61","model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench-Science 0.1 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"4f191cd5eb9caf57e48bc0e7","raw_value":59.0,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"59.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench-Science 0.1 (AA):Claude Opus 5.5"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench-Science:Terminal-Bench-Science 0.1 (AA)","source_record_id":"carbon-image:Terminal-Bench-Science 0.1 (AA):Claude Opus 5.5","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":59.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"b4428690505427f75d65ee7f08fbf5db","model_id":"m-cfe0652e9d685783","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Business workflows AutomationBench²; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":40.0,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"40.0%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row10:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"official-0dd3bed5c7dd:table0:row10:column1","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":40.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"35aa2ee19acc44897fc2654a03403f3a","model_id":"m-cfe0652e9d685783","original_version_label":"0.1","profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Agentic scientific research Terminal-Bench-Science 0.1³; preserve page footnotes and task-specific settings","protocol_id":"1975ca8c09f87d3fc7597c12","raw_value":58.7,"research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"58.7%","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row14:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench-Science:Agentic scientific research Terminal-Bench-Science 0.1","source_record_id":"official-0dd3bed5c7dd:table0:row14:column1","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":58.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"0e04d5cca5cf4605416149e9e081590e","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":66.4,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"66.4%¹","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row2:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-4e3ca8a95d36:table0:row2:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":66.4},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"2674db64c27edbe09263f61e9c5887af","model_id":"m-cfe0652e9d685783","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1846.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1846","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1846.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"56ec8a4eb3d885260337d38fdb4d7987","model_id":"m-cfe0652e9d685783","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1822.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1822","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1822.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","model_id":"m-cfe0652e9d685783","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"eff89ef5cfa18da15eddc05443bb6b8d","model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":55,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"037aa248f2b78a5c9d92b26e"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"037aa248f2b78a5c9d92b26e","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":55},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bab68b4a8cd2763b38ac011956432e06","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":65.15,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"65.15%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":65.15},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","model_id":"m-cfe0652e9d685783","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b69f2f163f3c7b09b02dfe152d26ca01","model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"215a23eaf57b6e179b6e1d675a2d74a4","model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.865895953757225,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5-xhigh:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.865895953757225},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"4c5a2ff45f0a237a692851070a84970c","model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.857225433526012,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.857225433526012},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ab181a64dcd14d2f58e0f2d48295da60","model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.858381502890173,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5-high:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.858381502890173},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e49cee0d44b11b8909b7ba5e93706c6d","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.95,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.95},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a58d05c58cb45d01767d253094a55889","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1759.61,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1759.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"282177141d1b28036a011934c6e4b8c6","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1620.14,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1620.14},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dd2f9e0cfcd8ae56db8e072077386bb9","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1682.43,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1682.43},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5e17d9b5f30d6a5c9321e7325441ffaa","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6333333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6333333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6be10aaf494f50e88c040d8d89850db8","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6553333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6553333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1846f249eb200a45fa8d28a4064825e1","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.639,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.639},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d1c630ff7e7d05f22ed35d2f38ae5c8","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6483333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6483333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"bfb4382ec51a22beabbc42c902403134","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3372727272727273,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372727272727273},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6f7a8c35429e18b671e356bb956a6ff1","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3500967117988395,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3500967117988395},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4a585623d13083bcceacf292b437dae","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33148661126500467,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33148661126500467},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e3f75977ffc03e709e7727cc60fb","model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3611374407582938,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3611374407582938},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f516e78ea47fddbbc174c19a1bf71c2c","model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5292537042934996,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5292537042934996},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9736e836ae76261125dd114b416c585d","model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6495283516401527,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6495283516401527},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bf4bd26804b45ca1abbcb387c34063a8","model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6121742919892179,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6121742919892179},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a3728f00fd503b759c0f1d5231e5c2e3","model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6320677483186213,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6320677483186213},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8c7b6663de4bc02f5ebccae026f35f62","model_id":"m-cfe0652e9d685783","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.59047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-opus-5-5:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59047619047619},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"68de8e9e106a5306483ea9f694b6328b","model_id":"m-cfe0652e9d685783","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.242857142857143,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-opus-5-5-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.242857142857143},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2df2e246ec0713b95677b8231fc92811","model_id":"m-cfe0652e9d685783","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.619047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-opus-5-5-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.619047619047619},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5ab0427c869adbf19d49c606c3e51010","model_id":"m-cfe0652e9d685783","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.433333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-opus-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.433333333333333},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"abdf9b74f43cb55343d6a9a8c60d8363","model_id":"m-cfe0652e9d685783","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.49047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-opus-5-5-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49047619047619},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","model_id":"m-cfe0652e9d685783","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"30b27abf5aed18095d3a9b6b94758349","model_id":"m-cfe0652e9d685783","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.158378282828283,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.158378282828283},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b8225447f72bc3d43b8b4532f9258364","model_id":"m-cfe0652e9d685783","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.437547474747475,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.437547474747475},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35121337a1bab95db38ef111b0d254fb","model_id":"m-cfe0652e9d685783","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.330793181818182,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.330793181818182},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"08da8ce2004becf17aefb666f1b20756","model_id":"m-cfe0652e9d685783","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.359278535353535,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.359278535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"fd85bfc6608f4e6c682dbcde61c416f0","model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.417,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.417},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2d951d0aa63880c5ff36a27d6635af8","model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.64,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.64},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eda897d41c6cfb5133a50e0ff16628d3","model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.514,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.514},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0676192c601bfd0fa6cb8d8c67ace8da","model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.556,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff13e1dbf1ca147eb2d8a88ccdde5448","model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.256,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.256},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1945008230269eaff07b848f5abccae0","model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.266,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.266},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9a4a7cf1a3bf9d3055807a16fdcba858","model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.256,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.256},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"613578548302526e230b9b56e0ad97cb","model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.288,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.288},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dfa5f82cc27ed08f1b2a7428eaa55225","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.313131313131313,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.313131313131313},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9974ff3635acb2a6ca5dd0bee3934ae","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"badc6e785a7434da9fa8c253a5cdf85d","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.525252525252525,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525252525252525},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"34b4b2f739837ed8d4e5f48d751ba4ef","model_id":"m-cfe0652e9d685783","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.565656565656566,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565656565656566},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"61fe9366db1e060fdb8fc46d0e9c624f","model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e5c48da134cbe5f70cebc4486243a9e9","model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.675,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.675},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4dacf01d99c21d74f50a08d4c77a29e","model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.54,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.54},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"be7409b17fd2bfac0bd622017a73969b","model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.595,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af85a0ab0efdfabaf0679421907beb34","model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1240.15,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1240.15},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"25e8f42755992fd7cd71ea55b139499d","model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1835.03,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1835.03},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63d4c8aedf3bee19f53391e2a2ca3339","model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1592.56,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1592.56},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca4f9e8d59e52ea2239db53073f5efb9","model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1713.6,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1713.6},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":61047,"numerator":5798875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1775},"family":"Terminal-Bench-Science","row_ids":["21af8993ff5f3e38ecc3caf66271b208"],"score":93.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":6200},"family":"HLE","row_ids":["3522855b1814574fd5b3733b32b77769"],"score":98.41},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4750},"family":"CritPt","row_ids":["46e4906d019b793c19b1d443cd35b178"],"score":93.14}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","peer_count":64,"percentile":98.41,"percentile_exact":{"denominator":63,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68}],"score":94.99,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":302841,"numerator":28338056},"families":[{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":100},"family":"GDPval-AA","row_ids":["b44123b055ba24f02daafca81efcb87b"],"score":100.0},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":96},"family":"Terminal-Bench","row_ids":["322131c4aa07af285df46f472061072d"],"score":96.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1750},"family":"AutomationBench","row_ids":["f100a726506576804779fdbd76208d6e"],"score":92.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7600},"family":"AA-Briefcase","row_ids":["5c16c97705a8f70907dc1ba381949448"],"score":98.7},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":250},"family":"OSWorld","row_ids":["cb71933d15a49daa0088c6b64854af7f"],"score":83.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2100},"family":"AA-AnalystAgent","row_ids":["1a02ab6144866fd1b960f19d8fd21513"],"score":91.3}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","peer_count":82,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"4.0","peer_count":76,"percentile":96.0,"percentile_exact":{"denominator":1,"numerator":96},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":78,"percentile":98.7,"percentile_exact":{"denominator":77,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.1","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","peer_count":24,"percentile":91.3,"percentile_exact":{"denominator":23,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625}],"score":93.57,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":1275},"families":[{"benchmark_version":"2.0","exact":{"denominator":1,"numerator":100},"family":"FrontierCode","row_ids":["a1969d9eb66e2bed1a98e36808c17360"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1250},"family":"Vals Code Migration","row_ids":["d2e15e2796cf44b3aabbae2da4bd5b3f"],"score":96.15}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.0","peer_count":50,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"1.1","peer_count":27,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01}],"score":98.08,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2822875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3425},"family":"GDP.pdf","row_ids":["77cd16d89e49025a3285c182b2da553d"],"score":90.13},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1550},"family":"SimpleQA Verified","row_ids":["144b27b350bf50d4c19f78baa94c183a"],"score":91.18},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3500},"family":"AA-Omniscience","row_ids":["3247fbdae3f488e2395593b52951cc4b","035a4d676bedc8d21713e0b6aa5bc0ad"],"score":66.04}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227}],"score":82.45,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":459,"numerator":37630},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"Humanity Sixth Sense","row_ids":["78db71948ffd6a9f7a24a59779294d90"],"score":88.89},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1650},"family":"MMMU-Pro","row_ids":["3dabe34a660554a2d248f440d9ae144c"],"score":97.06},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":60},"family":"Chartography","row_ids":["d2e4430a89fa4e3972bc81f62bbdad07"],"score":60.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","peer_count":28,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":60.0,"percentile_exact":{"denominator":1,"numerator":60},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4}],"score":81.98,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","peer_count":64,"percentile":98.41,"percentile_exact":{"denominator":63,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","peer_count":82,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"4.0","peer_count":76,"percentile":96.0,"percentile_exact":{"denominator":1,"numerator":96},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":78,"percentile":98.7,"percentile_exact":{"denominator":77,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.1","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","peer_count":24,"percentile":91.3,"percentile_exact":{"denominator":23,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.0","peer_count":50,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"1.1","peer_count":27,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","peer_count":28,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":60.0,"percentile_exact":{"denominator":1,"numerator":60},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4}],"result_updated_at":"2026-09-28","score":91.44,"score_exact":{"denominator":42566119596,"numerator":3892435594141}}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"source_id":"anthropic/claude-opus-5-5"},"m-d031501e2d0117d8":{"id":"m-d031501e2d0117d8","in_window":true,"name":"Apodex 1.1","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","model_id":"m-d031501e2d0117d8","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","model_id":"m-d031501e2d0117d8","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-30","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":32214,"numerator":1462025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3600},"family":"GPQA Diamond","row_ids":["90978c171b499bb2efc92c2da0d369b4"],"score":61.02},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["5952fe944fd804ef75ab5466dc485304"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":250},"family":"HLE","row_ids":["7efb7f319954edefb37b62ce3d3ff826"],"score":35.71}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":61.02,"percentile_exact":{"denominator":59,"numerator":3600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116}],"score":45.38,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":4621617,"numerator":208057310},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"APEX-Agents","row_ids":["b5b0e9a07c5a8ced9fdf4c9ac74df7fc"],"score":50.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":775},"family":"tau3-Banking","row_ids":["84d4042d1c82bfee0ac491e53b4fa917"],"score":59.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4400},"family":"GDPval-AA","row_ids":["5a309e2224372571fd72b2e3de2fe6e1"],"score":54.32},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["cafc26cd33f16c1828c33bad26ea21a7"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":925},"family":"AutomationBench","row_ids":["2d900bebd8be5f37326880467236361d"],"score":48.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3400},"family":"AA-Briefcase","row_ids":["8d8e39fb86f2781e20fa7a7f68dc2ca4"],"score":44.16}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":59.62,"percentile_exact":{"denominator":13,"numerator":775},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","peer_count":82,"percentile":54.32,"percentile_exact":{"denominator":81,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":48.68,"percentile_exact":{"denominator":19,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","peer_count":78,"percentile":44.16,"percentile_exact":{"denominator":77,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16}],"score":45.02,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":38,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["ec9b7d6c1e90c4addeb4e33b7ca0403e"],"score":48.03}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"score":48.03,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":950},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":950},"family":"MMMU-Pro","row_ids":["d16dcb5ce825969190ea4f39de9f84b5"],"score":55.88}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867}],"score":55.88,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":61.02,"percentile_exact":{"denominator":59,"numerator":3600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":59.62,"percentile_exact":{"denominator":13,"numerator":775},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","peer_count":82,"percentile":54.32,"percentile_exact":{"denominator":81,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":48.68,"percentile_exact":{"denominator":19,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","peer_count":78,"percentile":44.16,"percentile_exact":{"denominator":77,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867}],"result_updated_at":null,"score":47.24,"score_exact":{"denominator":148335419232,"numerator":7007785076525}}],"release_date":"2026-08-30","release_status":"published","release_verified":true,"source_id":"published/aa:apodex-1-1"},"m-d0874ed91fd69e8e":{"id":"m-d0874ed91fd69e8e","in_window":true,"name":"LFM2.5-2.6B","observed_records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1},{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"539fe5c7d8f8256e0c95c73bbfd93eb2","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.0,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"59.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":59.0},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"7b398827aeb7745497533eabe61541f6","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":76.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"76.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":76.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA-Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c750a2ef43ab55535245679f96847a15","model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA-Diamond; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"62f4969756e5cb55e0903043","raw_value":55.8,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"55.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row22:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA-Diamond","source_record_id":"twelve-minicpm2-card:table0:row22:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":55.8},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"79e538fa5b2bc837bbbc3882037e1347","model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":61.1,"research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"61.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":61.1},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBenchv6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"coding","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"6b837ab1aea3633e64a6d8b789a444d0","model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","profile_key":"native","protocol_description":"Official published comparison; LiveCodeBenchv6; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"7115ec33f7a119b1fc3bb499","raw_value":59.41,"research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.41","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBenchv6","source_record_id":"twelve-liquid26-card:table4:row5:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.41},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88},{"benchmark_family":"ToolSandbox","benchmark_id":"b-bbde7565ae63d4bedf3a","benchmark_label":"ToolSandbox","benchmark_version":"current maintainer task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"4f42e2391e72e7491dec4c044d47ebdc","model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Apple task suite; evaluator protocol separate","profile_key":"native","protocol_description":"Official published comparison; ToolSandbox; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"82db2dbdf1f340057160a4b5","raw_value":77.83,"research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"77.83","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row10:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:ToolSandbox:ToolSandbox","source_record_id":"twelve-liquid26-card:table4:row10:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":77.83},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-04","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":177,"numerator":500},"families":[{"benchmark_version":"2026","exact":{"denominator":1,"numerator":0},"family":"AIME","row_ids":["1aa7620d5715919ff1d95bb835bf5bb1"],"score":0.0},{"benchmark_version":"500-item subset","exact":{"denominator":1,"numerator":0},"family":"MATH-500","row_ids":["d2a6fbf18a7d28ad26aab8f6363bd884"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":500},"family":"GPQA Diamond","row_ids":["c0382c8dd0c506272b3fef2eaf1a7bcf"],"score":8.47}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":8.47,"percentile_exact":{"denominator":59,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758}],"score":2.82,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3159,"numerator":120050},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":150},"family":"tau3-Banking","row_ids":["ad5761dce64ad5d75bd8fe071f2c9120"],"score":11.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":200},"family":"GDPval-AA","row_ids":["7d56dec4d055d3d9b1237e11f0bf7ef4"],"score":2.47},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["78759a40401be7663e3957b80c649826"],"score":100.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","peer_count":82,"percentile":2.47,"percentile_exact":{"denominator":81,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88}],"score":38.0,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"release_v6","exact":{"denominator":3,"numerator":100},"family":"LiveCodeBench","row_ids":["ff54ed0b44635ad5d8a708f3d17d4063"],"score":33.33}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1}],"score":33.33,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":36,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"IFBench","row_ids":["49b7d39b3c4a3eaca06d3883d0335282"],"score":11.11},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"Multi-IF","row_ids":["ddcaace3bb0b4a1452f5dea314dd058c"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":200},"family":"IFEval","row_ids":["a24e132d4558f677a1ce0a399471dbb9"],"score":66.67},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":25},"family":"MMLU-Pro","row_ids":["239f2b6a72085e3eb63d599375b5b516"],"score":25.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2}],"score":50.69,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":8.47,"percentile_exact":{"denominator":59,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","peer_count":82,"percentile":2.47,"percentile_exact":{"denominator":81,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2}],"result_updated_at":null,"score":30.01,"score_exact":{"denominator":3354858,"numerator":100691975}}],"release_date":"2026-08-04","release_status":"published","release_verified":true,"source_id":"published/aa:lfm2-5-2-6b"},"m-d123610543e5f5ed":{"id":"m-d123610543e5f5ed","in_window":true,"name":"Claude Sonnet 5.5","observed_records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4ea887aa7cc58c78a364e0a2b65d69bf","model_id":"m-d123610543e5f5ed","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1840.0,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"1840","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row2:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-cdb1d51fae04:table0:row2:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"分","value":1840.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"be4f3c7647faa1a11f91719e70b95e94","model_id":"m-d123610543e5f5ed","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Knowledge work AA-Briefcase v1.1; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1824.0,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"1824","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row4:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-cdb1d51fae04:table0:row4:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":1824.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","model_id":"m-d123610543e5f5ed","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"2b76f8801c1b2fa6370b3cd152e2d18c","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":70.6,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"70.6%","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row11:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-cdb1d51fae04:table0:row11:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":70.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"aa4d2045301c956c92f199a1137acd1b","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":70.6,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"70.6%","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row2:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-4e3ca8a95d36:table0:row2:column1","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":70.6},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"197a72fcf9c4e9c50a6d47d1363590c9","model_id":"m-d123610543e5f5ed","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1844.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1844","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column1","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1844.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"5584934903b9f42c4853160cdcdfd33e","model_id":"m-d123610543e5f5ed","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1811.0,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1811","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column1","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1811.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"651c1df0fae8b1a615b85f4f0be4accd","model_id":"m-d123610543e5f5ed","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.1,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"80.1% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column1","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":80.1},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"91fafd2de745057ad9bf6ab81bdb8068","model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":61.6,"research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"61.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column1","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":61.6},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"886c25f56626744a389c5cae640c1a98","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":64.14,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"64.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":64.14},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","model_id":"m-d123610543e5f5ed","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"92ec36c88ee8d600291d69ec7635ffff","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1439.22,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1439.22},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb1c409847579e51aedffba4198d3f32","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1634.38,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1634.38},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3b77708ce06f7b5a6cb99e9490c8aaab","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1741.64,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1741.64},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c17eb7418f879fc7726a28c22e5f625","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1267.6,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1267.6},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0736adb38c9f1b0b0e5f9c5e8a6ca937","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45766666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"128e04e09b3fabd17220678e32e72ae3","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5033333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"66cc769d9bd339411213561bcf30204a","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.509,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.509},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5b0659e82dcbfbce72e59157d76fb85e","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.46,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.46},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f8faab2a2e7ddf8942d575b1929178d0","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49969268592501537,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49969268592501537},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f3e302f89350da1b21e2f07e2094a484","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.363758389261745,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.363758389261745},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8a40625f029e02fdf5c9f92e257edff","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38289205702647655,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38289205702647655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a2bd8bd3e7de3e0e4b3b1488063cb3b","model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5234567901234568,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5234567901234568},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4e9713a7ebe95a74718df8d8f4c64280","model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5490962996121462,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5490962996121462},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b13a99ba92468133c06363b61c90186e","model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5942923026067375,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5942923026067375},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d03a69fe023785fce36542500eaba59","model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6552026949665557,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6552026949665557},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1cd8cc051939d915fc3cf0f59cd83","model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4942092641344955,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4942092641344955},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"86553476b341937aa8eee07e46a4463a","model_id":"m-d123610543e5f5ed","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.161904761904762,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.161904761904762},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e461f0a0d23ed77d65b8c3f2496dafc2","model_id":"m-d123610543e5f5ed","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"36967061bad853aca1202fc303de74e6","model_id":"m-d123610543e5f5ed","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.523809523809524,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.523809523809524},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8219d3f5fb2653720b3130ec56d5201c","model_id":"m-d123610543e5f5ed","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.533333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.533333333333333},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"41202251510cd2e6bd0137c69d1ea82f","model_id":"m-d123610543e5f5ed","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.104761904761905,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.104761904761905},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0b57e5c29cd8d8cb074738b9847e0498","model_id":"m-d123610543e5f5ed","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.109708080808081,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.109708080808081},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cac52d356083ab03e3854da85f8b35ba","model_id":"m-d123610543e5f5ed","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257380050505051,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257380050505051},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a83a17e51454694dde520f7313db4bbd","model_id":"m-d123610543e5f5ed","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.356494696969697,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356494696969697},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","model_id":"m-d123610543e5f5ed","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a0695f474729b388d8f10af74cb716d8","model_id":"m-d123610543e5f5ed","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0712575757575758,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0712575757575758},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eb06088be6a3a408b8ad1fa6b778d439","model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.238,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.238},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"112095e8d3e60e4f325854f0f184fb52","model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.338,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.338},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c9c23f8fe04233a049af6bc649777d82","model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.441,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.441},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c57acd914b7734768f36308b43cd96b9","model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.533,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.533},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ddc070bb6f174f9dc5abf9813b7381e5","model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.206,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.206},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24fbe662c8a1ef8b56fdcfe0e0f880b0","model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be61fe78354f92c5a09f9fd12b846a6b","model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fd701c14e4a8146437b2c47fa67f66b3","model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.246,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.246},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"70ea9992b63f49842a0088a5671186b5","model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.16,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"691583ddf4d66c67c5d9ae0659216979","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0f8e9ea1952ecec8eee01ea2f9fb834c","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.439393939393939,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.439393939393939},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"99ee05d8dd3d34210b8903dd91db2a89","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"38befbded575e4ae84bfa33192ba6294","model_id":"m-d123610543e5f5ed","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.207070707070707,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.207070707070707},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7a29bfd2327e904977910e6c3c6d954c","model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"86505250c8794e191bff51e86f1c963c","model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.515,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.515},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6f0c021a8bd1d9a1648510ee3b10e24b","model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.575,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cf09052620c016415aaa84efe5595fe5","model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.26,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Medium, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7c053e5e28c9e886b3b416fc4b668ea1","model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1328.63,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1328.63},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (High, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"704953eec5d98ae5b957449f85bedf0f","model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1556.89,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1556.89},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Xhigh, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f65424ef6863771650ecf15f98971e38","model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1735.08,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1735.08},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Low, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6df0fc05dc37bf7c480d8b5eddf51dc2","model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1181.72,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1181.72},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":46512,"numerator":3838925},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":700},"family":"FrontierMath","row_ids":["a0152968ddf4e79954f370988a3aa877"],"score":77.78},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1775},"family":"Terminal-Bench-Science","row_ids":["0fd84fb90c507070cc93b2671c1dacd9"],"score":93.42},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":275},"family":"HLE","row_ids":["36082ffe886fbeb2647876856cf1e5c0"],"score":68.75},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4600},"family":"CritPt","row_ids":["57f29ac78c3749b38fa2eef65ef43c04"],"score":90.2}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","peer_count":52,"percentile":90.2,"percentile_exact":{"denominator":51,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59}],"score":82.54,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":35397,"numerator":3458243},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2600},"family":"GDPval-AA","row_ids":["c0ede537ba74a003e6de650e7903fd9e"],"score":96.3},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":292},"family":"Terminal-Bench","row_ids":["620eabc34583900c0a5610d2452e8c0e"],"score":97.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1800},"family":"AutomationBench","row_ids":["7e4dabafdeb0615d7fe8c2c4159042bd"],"score":94.74},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"AA-Briefcase","row_ids":["609c8649b3a04b44b0d6807fdfe15427"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":100},"family":"OSWorld","row_ids":["e0c59ec207d44193213f25c93cf18f5b"],"score":100.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2250},"family":"AA-AnalystAgent","row_ids":["b14e1b2737d16e7bb80a947abb8a320a"],"score":97.83}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","peer_count":82,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"4.0","peer_count":76,"percentile":97.33,"percentile_exact":{"denominator":3,"numerator":292},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":78,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"score":97.7,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":49,"numerator":4750},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4600},"family":"FrontierCode","row_ids":["2bdac0e90d63fa507e78af95b1c832a5"],"score":93.88},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"Vals Code Migration","row_ids":["165b56e6a0826f1ed0dc82b37b0885fe"],"score":100.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.0","peer_count":50,"percentile":93.88,"percentile_exact":{"denominator":49,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"1.1","peer_count":27,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88}],"score":96.94,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":3278075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1675},"family":"GDP.pdf","row_ids":["7992d55958a238d712427a450a500af9"],"score":88.16},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":650},"family":"SimpleQA Verified","row_ids":["b5dfd8ffb656e8526bed64b86b168f2a"],"score":38.24},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3450},"family":"AA-Omniscience","row_ids":["911853403e469f557e0c742924a5bacb","553708168125ea365f4f1a89df474d93"],"score":65.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923}],"score":63.83,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["34ec823cb4bb85ff7965822a48a3f4ce"],"score":100.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6}],"score":100.0,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","peer_count":52,"percentile":90.2,"percentile_exact":{"denominator":51,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","peer_count":82,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"4.0","peer_count":76,"percentile":97.33,"percentile_exact":{"denominator":3,"numerator":292},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":78,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.0","peer_count":50,"percentile":93.88,"percentile_exact":{"denominator":49,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"1.1","peer_count":27,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6}],"result_updated_at":"2026-10-07","score":87.21,"score_exact":{"denominator":100015497792,"numerator":8722585306847}}],"release_date":"2026-09-28","release_status":"published","release_verified":true,"source_id":"anthropic/claude-sonnet-5-5"},"m-d23e83fbb1d3aabc":{"id":"m-d23e83fbb1d3aabc","in_window":true,"name":"Hy3","observed_records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"8a7def0b05a244600845308f3acf91c6","model_id":"m-d23e83fbb1d3aabc","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":90.9,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"90.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":90.9},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","model_id":"m-d23e83fbb1d3aabc","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"ea9839b4879e46422a95dd0f7e029b6a","model_id":"m-d23e83fbb1d3aabc","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":16.1,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"16.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":16.1},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","model_id":"m-d23e83fbb1d3aabc","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":413,"numerator":23125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4500},"family":"GPQA Diamond","row_ids":["9b6a2e4e945c90d8754fc862b2e6383b"],"score":76.27},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":250},"family":"HLE","row_ids":["a0be97a286389603a66382ac4d794e81"],"score":35.71}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":76.27,"percentile_exact":{"denominator":59,"numerator":4500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116}],"score":55.99,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":1320462,"numerator":43715939},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["869a19a4a7bb34031859441309a0c7b5"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":700},"family":"tau3-Banking","row_ids":["a00dadd2de4cef1d6ffa67d4028686b0"],"score":53.85},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3100},"family":"GDPval-AA","row_ids":["3e2d18ecd2bf6e50f2e15788da94966d"],"score":38.27},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["87cb9fb97ee59be625938b752ce0bb36"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":725},"family":"AutomationBench","row_ids":["df8bf63c05795b45b4fd82cd57c8b504"],"score":38.16},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":400},"family":"AA-Briefcase","row_ids":["a684e0a09f0e331ba14d29191ea44f68"],"score":36.36}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","peer_count":82,"percentile":38.27,"percentile_exact":{"denominator":81,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","peer_count":78,"percentile":36.36,"percentile_exact":{"denominator":11,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24}],"score":33.11,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["6f7ac5ece00998e78042608639ac6e80"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["2081b40577b770d6210bb3099918429f"],"score":48.03}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"score":48.03,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":76.27,"percentile_exact":{"denominator":59,"numerator":4500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","peer_count":82,"percentile":38.27,"percentile_exact":{"denominator":81,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","peer_count":78,"percentile":36.36,"percentile_exact":{"denominator":11,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"result_updated_at":null,"score":35.42,"score_exact":{"denominator":9816314508,"numerator":347716592885}}],"release_date":"2026-07-06","release_status":"published","release_verified":true,"source_id":"tencent/hy3"},"m-d2febb93118535a2":{"id":"m-d2febb93118535a2","in_window":true,"name":"K2 Horizon 375B A23B","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","model_id":"m-d2febb93118535a2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","model_id":"m-d2febb93118535a2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","model_id":"m-d2febb93118535a2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","model_id":"m-d2febb93118535a2","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","model_id":"m-d2febb93118535a2","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","model_id":"m-d2febb93118535a2","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","model_id":"m-d2febb93118535a2","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","model_id":"m-d2febb93118535a2","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":115841125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3900},"family":"GPQA Diamond","row_ids":["b588de5f182f53797aeaba57de63e608"],"score":66.1},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["2f6e14dbb9becc8bf0feb78f2973083f"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1900},"family":"HLE","row_ids":["03d73659cb7b3abbbd4921c8546a8418"],"score":30.16},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1400},"family":"CritPt","row_ids":["5f0a8e1454688084cb0e096162e1eff2"],"score":27.45}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":66.1,"percentile_exact":{"denominator":59,"numerator":3900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond","peer_count":64,"percentile":30.16,"percentile_exact":{"denominator":63,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","peer_count":52,"percentile":27.45,"percentile_exact":{"denominator":51,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12}],"score":35.25,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":446904872},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":900},"family":"tau3-Banking","row_ids":["348ef206dd0fa625708f257a200455f5"],"score":69.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4900},"family":"GDPval-AA","row_ids":["3637cd472b3e658153b3601c1c953f3e"],"score":60.49},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":48},"family":"Terminal-Bench","row_ids":["612528520707bc7ebd5f35810f7efeb6"],"score":48.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1000},"family":"AutomationBench","row_ids":["d30d4db6bfda99c8317f5eddf5d0e239"],"score":52.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4600},"family":"AA-Briefcase","row_ids":["4c2ed1cac020a530bd05f7d661a8c39c"],"score":59.74}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2.1","peer_count":82,"percentile":60.49,"percentile_exact":{"denominator":81,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"4.0","peer_count":76,"percentile":48.0,"percentile_exact":{"denominator":1,"numerator":48},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":78,"percentile":59.74,"percentile_exact":{"denominator":77,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48}],"score":58.02,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1900},"family":"FrontierCode","row_ids":["b9e8a0463a865f6221caaaf1d3aecc9d"],"score":38.78}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"2.0","peer_count":50,"percentile":38.78,"percentile_exact":{"denominator":49,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465}],"score":38.78,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1007,"numerator":40025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":650},"family":"GDP.pdf","row_ids":["65f6a64cea971ea25fa466fbee3340a6"],"score":34.21},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2400},"family":"AA-Omniscience","row_ids":["d414f5fe5dd8b0e7a11df833233b1edf","7d19cde6db2f8d3d5e0cb451c988a92a"],"score":45.28}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506}],"score":39.75,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":66.1,"percentile_exact":{"denominator":59,"numerator":3900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond","peer_count":64,"percentile":30.16,"percentile_exact":{"denominator":63,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","peer_count":52,"percentile":27.45,"percentile_exact":{"denominator":51,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2.1","peer_count":82,"percentile":60.49,"percentile_exact":{"denominator":81,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"4.0","peer_count":76,"percentile":48.0,"percentile_exact":{"denominator":1,"numerator":48},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":78,"percentile":59.74,"percentile_exact":{"denominator":77,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"2.0","peer_count":50,"percentile":38.78,"percentile_exact":{"denominator":49,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506}],"result_updated_at":null,"score":43.36,"score_exact":{"denominator":41274330401304,"numerator":1789607818257619}}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"source_id":"published/aa:k2-horizon-375b-a23b"},"m-d48cde87e1f9c28d":{"id":"m-d48cde87e1f9c28d","in_window":true,"name":"A.X-K2","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["6d2464487e35f420dae74891c6e0dec0"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":513513,"numerator":11596703},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":400},"family":"tau3-Banking","row_ids":["f6ef999603d2cd2e7f3ecd5e4118aa3e"],"score":30.77},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":800},"family":"GDPval-AA","row_ids":["518222bbae6a6938c2fc68e86213e044"],"score":29.63},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["46a72c291abf86482bd30a1de9ec49ac"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":325},"family":"AutomationBench","row_ids":["3e7bd955b2670c77e55665a0183fa94c"],"score":17.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1700},"family":"AA-Briefcase","row_ids":["628eb864f68ddd810e173dc996c88c01"],"score":22.08}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","peer_count":82,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":17.11,"percentile_exact":{"denominator":19,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","peer_count":78,"percentile":22.08,"percentile_exact":{"denominator":77,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17}],"score":22.58,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1175},"family":"GDP.pdf","row_ids":["10af516d49f62278a451fea09c7878c1"],"score":30.92}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"score":30.92,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","peer_count":82,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":17.11,"percentile_exact":{"denominator":19,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","peer_count":78,"percentile":22.08,"percentile_exact":{"denominator":77,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"result_updated_at":null,"score":36.88,"score_exact":{"denominator":424161738,"numerator":15641285435}}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"source_id":"published/aa:a-x-k2"},"m-da421acad97d607b":{"id":"m-da421acad97d607b","in_window":true,"name":"MiniMax-M3","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","model_id":"m-da421acad97d607b","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","model_id":"m-da421acad97d607b","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","model_id":"m-da421acad97d607b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","model_id":"m-da421acad97d607b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","model_id":"m-da421acad97d607b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","model_id":"m-da421acad97d607b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","model_id":"m-da421acad97d607b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","model_id":"m-da421acad97d607b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ab24b7c89863efc898bb5b0dc33b7ea8","model_id":"m-da421acad97d607b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0202020202020202,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0202020202020202},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","model_id":"m-da421acad97d607b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","model_id":"m-da421acad97d607b","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":730184,"numerator":43081725},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5500},"family":"GPQA Diamond","row_ids":["67157f1183905451cf4dded1391fab8e"],"score":93.22},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["1a6a33cd4cc454e60378e0a0694263f3"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1100},"family":"HLE","row_ids":["4912807a823c5e919553ca48ade2b67a"],"score":52.38},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2600},"family":"CritPt","row_ids":["3d5749511482d0fbdaddfa896055ac6a"],"score":50.98}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":93.22,"percentile_exact":{"denominator":59,"numerator":5500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond","peer_count":64,"percentile":52.38,"percentile_exact":{"denominator":21,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","peer_count":52,"percentile":50.98,"percentile_exact":{"denominator":51,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28}],"score":59.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":992107116,"numerator":27805451125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":375},"family":"tau3-Banking","row_ids":["0494d5d3e701ac0729bd55524064b2fa"],"score":28.85},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4600},"family":"GDPval-AA","row_ids":["7ee504c210f7e12209fb9a5011615203"],"score":56.79},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":25},"family":"Terminal-Bench","row_ids":["8f361f3f1d062dfe3e449864fa84aec9"],"score":6.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":775},"family":"AutomationBench","row_ids":["c964aa700a761259aa3ab0adef555cba"],"score":40.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3700},"family":"AA-Briefcase","row_ids":["703518382f29bb0a80ddba827b81e22d"],"score":48.05},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["df3f56d8282f0ce2c7d297df530d227a"],"score":11.11},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":100},"family":"AA-AnalystAgent","row_ids":["8cef68049263078fc3646fc52bd5e75a"],"score":4.35}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","included":true,"model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":28.85,"percentile_exact":{"denominator":13,"numerator":375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2.1","peer_count":82,"percentile":56.79,"percentile_exact":{"denominator":81,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","included":true,"model_id":"m-da421acad97d607b","original_version_label":"4.0","peer_count":49,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","included":true,"model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":78,"percentile":48.05,"percentile_exact":{"denominator":77,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA run","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","peer_count":24,"percentile":4.35,"percentile_exact":{"denominator":23,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"score":28.03,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":21950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3000},"family":"FrontierCode","row_ids":["3f103c270013e8e3a189c86aaaac75c3"],"score":61.22},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":100},"family":"Vals Code Migration","row_ids":["fbec8a823687d43757fc0efa8ad26423"],"score":7.69}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"2.0","peer_count":50,"percentile":61.22,"percentile_exact":{"denominator":49,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","included":true,"model_id":"m-da421acad97d607b","original_version_label":"1.1","peer_count":27,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9}],"score":34.46,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":3021,"numerator":181675},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["89b2619aee2f39a13fb5f1677cbe7c28"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":775},"family":"GDP.pdf","row_ids":["7e0fa158109c3646bd13a80e33ec79ee"],"score":40.79},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["253af28f0e697d86e6b8e62f78eae062","e14051d6dbdd6fd61ec4748398bf8936"],"score":39.62}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043}],"score":60.14,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":750},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":750},"family":"MMMU-Pro","row_ids":["9161d7aea787f48463e1dc1ca6952542"],"score":44.12}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977}],"score":44.12,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":93.22,"percentile_exact":{"denominator":59,"numerator":5500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond","peer_count":64,"percentile":52.38,"percentile_exact":{"denominator":21,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","peer_count":52,"percentile":50.98,"percentile_exact":{"denominator":51,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","included":true,"model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":28.85,"percentile_exact":{"denominator":13,"numerator":375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2.1","peer_count":82,"percentile":56.79,"percentile_exact":{"denominator":81,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","included":true,"model_id":"m-da421acad97d607b","original_version_label":"4.0","peer_count":49,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","included":true,"model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":78,"percentile":48.05,"percentile_exact":{"denominator":77,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA run","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","peer_count":24,"percentile":4.35,"percentile_exact":{"denominator":23,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"2.0","peer_count":50,"percentile":61.22,"percentile_exact":{"denominator":49,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","included":true,"model_id":"m-da421acad97d607b","original_version_label":"1.1","peer_count":27,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977}],"result_updated_at":null,"score":45.09,"score_exact":{"denominator":32455029033504,"numerator":1463328516402595}}],"release_date":"2026-06-01","release_status":"published","release_verified":true,"source_id":"minimax/MiniMax-M3"},"m-dabee5e3d87d8a7a":{"id":"m-dabee5e3d87d8a7a","in_window":true,"name":"MiMo-V2.6-Pro","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c254195ad65c522fa9609ac4a64d415d","model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7be2038bfeb07e68eff4c9e5fd86646","model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.348484848484849,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mimo-v2-6-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348484848484849},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375}],"official_evidence":{"artifact_sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","basis":"official_release_page","date":"2026-09-22","sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","url":"https://mimo.mi.com/docs/en-US/updates/model"},"primary_profile_id":"native","profiles":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":3591,"numerator":160750},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":350},"family":"Terminal-Bench-Science","row_ids":["17afba9dea11c1bfb4fcda8f7ceb8fe4"],"score":18.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3100},"family":"HLE","row_ids":["2e68084b0515fdc6a19fe2b7a59814d2"],"score":49.21},{"benchmark_version":"2 Hard","exact":{"denominator":3,"numerator":200},"family":"CritPt","row_ids":["2f0eebfa718074ef0cab557ff07d209b"],"score":66.67}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","peer_count":64,"percentile":49.21,"percentile_exact":{"denominator":63,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","peer_count":52,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34}],"score":44.76,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":605682,"numerator":40432325},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":800},"family":"GDPval-AA","row_ids":["82bb5fcac208ae67d3e4a7d3de124d5d"],"score":88.89},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":325},"family":"Terminal-Bench","row_ids":["65da79b527ae8b650ca14c9d0ac7489d"],"score":54.17},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1375},"family":"AutomationBench","row_ids":["45f30ef9c9577f4abce574fa0b12543a"],"score":72.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6100},"family":"AA-Briefcase","row_ids":["c6fb2c23c770d5e7bda4e380c2517fe7"],"score":79.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":900},"family":"AA-AnalystAgent","row_ids":["59ce77c04181c398c36c852841fb4ecf"],"score":39.13}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","peer_count":82,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","peer_count":49,"percentile":54.17,"percentile_exact":{"denominator":6,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":78,"percentile":79.22,"percentile_exact":{"denominator":77,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","peer_count":24,"percentile":39.13,"percentile_exact":{"denominator":23,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375}],"score":66.76,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":34275},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3200},"family":"FrontierCode","row_ids":["f7aceff949a2ebec04ba4fe1d1d7e419"],"score":65.31},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":550},"family":"Vals Code Migration","row_ids":["c811b7493c2ca32eedcc762a5f354d58"],"score":42.31}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","peer_count":50,"percentile":65.31,"percentile_exact":{"denominator":49,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","peer_count":27,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79}],"score":53.81,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":58000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1400},"family":"GDP.pdf","row_ids":["0c01c6860fffc32ef81ab819ece21423"],"score":73.68},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2200},"family":"AA-Omniscience","row_ids":["d99f9179103c8156a9cedc7180b05c42","b3118336d86ca34d1fd6619d5a595882"],"score":41.51}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376}],"score":57.6,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":200},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["73994f722364016b18f62a10e6695b38"],"score":22.22}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","peer_count":28,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2}],"score":22.22,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","peer_count":64,"percentile":49.21,"percentile_exact":{"denominator":63,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","peer_count":52,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","peer_count":82,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","peer_count":49,"percentile":54.17,"percentile_exact":{"denominator":6,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":78,"percentile":79.22,"percentile_exact":{"denominator":77,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","peer_count":24,"percentile":39.13,"percentile_exact":{"denominator":23,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","peer_count":50,"percentile":65.31,"percentile_exact":{"denominator":49,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","peer_count":27,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","peer_count":28,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2}],"result_updated_at":null,"score":52.38,"score_exact":{"denominator":35054451432,"numerator":1836253825765}}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"source_id":"xiaomi/mimo-v2.6-pro"},"m-db208dec31031420":{"id":"m-db208dec31031420","in_window":true,"name":"Muse Spark Open","observed_records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","model_id":"m-db208dec31031420","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","model_id":"m-db208dec31031420","original_version_label":"Diamond","profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark Open / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8362f29198914b33fdad64ba36fb745e","model_id":"m-db208dec31031420","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.57,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.57%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row46"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row46","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.57},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","model_id":"m-db208dec31031420","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","model_id":"m-db208dec31031420","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","model_id":"m-db208dec31031420","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","model_id":"m-db208dec31031420","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","model_id":"m-db208dec31031420","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","model_id":"m-db208dec31031420","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","model_id":"m-db208dec31031420","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440f5ca7aab0d03094cae5ee8558ca3d","model_id":"m-db208dec31031420","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.263,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-spark-open:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","model_id":"m-db208dec31031420","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","model_id":"m-db208dec31031420","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","model_id":"m-db208dec31031420","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","model_id":"m-db208dec31031420","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":469404,"numerator":20872015},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":200},"family":"FrontierMath","row_ids":["c2edb36eb77a36c2b7e3b5065aafdd4c"],"score":22.22},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4000},"family":"GPQA Diamond","row_ids":["bf4b9ff10398e116bbd0ed6eb4a0a88d"],"score":67.8},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["356d1ac12dc91c2d1cadb615b7e3023b"],"score":46.15},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":125},"family":"HLE","row_ids":["04518aae51e855cbdd25679b26d57f51"],"score":31.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["3eef34fe8a68f3f65cb23f07be80c5e4"],"score":54.9}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","included":true,"model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":67.8,"percentile_exact":{"denominator":59,"numerator":4000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","included":true,"model_id":"m-db208dec31031420","original_version_label":"0.1","peer_count":53,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":44.46,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":13945607},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5000},"family":"GDPval-AA","row_ids":["c140b9f2f3b8b047f022d22a6670f4c4"],"score":61.73},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":164},"family":"Terminal-Bench","row_ids":["6d1fe83c673cc3e53108a1f798f9857e"],"score":54.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1150},"family":"AutomationBench","row_ids":["cc297748bec17bb0b4aa16e861980849"],"score":60.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4500},"family":"AA-Briefcase","row_ids":["00f62480bdc3ab7dea00bbda96ff9331"],"score":58.44}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2.1","peer_count":82,"percentile":61.73,"percentile_exact":{"denominator":81,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","included":true,"model_id":"m-db208dec31031420","original_version_label":"4.0","peer_count":76,"percentile":54.67,"percentile_exact":{"denominator":3,"numerator":164},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","included":true,"model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":78,"percentile":58.44,"percentile_exact":{"denominator":77,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69}],"score":58.84,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":15950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1700},"family":"FrontierCode","row_ids":["ca9b3bebd5cf2b1f50a3f98180ff0076"],"score":34.69},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":200},"family":"Vals Code Migration","row_ids":["aeba1081b4abe0557cd65f41eca5f8b3"],"score":15.38}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","included":true,"model_id":"m-db208dec31031420","original_version_label":"2.0","peer_count":50,"percentile":34.69,"percentile_exact":{"denominator":49,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","included":true,"model_id":"m-db208dec31031420","original_version_label":"1.1","peer_count":27,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89}],"score":25.04,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":8496475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1375},"family":"GDP.pdf","row_ids":["7048e36b711cfdf67298d82bfa3d5f86"],"score":72.37},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1100},"family":"SimpleQA Verified","row_ids":["2fc7380a8d1e0cbfdc6837142943acdd"],"score":64.71},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"MultiChallenge","row_ids":["fc1a81b2311badda3a38cf9a6a28b5f4"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3900},"family":"AA-Omniscience","row_ids":["dbcce4096cefcfd79cea60ad99bf1516","0b6767461c06c4b95e61401653221d25"],"score":73.58}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","included":true,"model_id":"m-db208dec31031420","original_version_label":"current AA protocol","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","included":true,"model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","included":true,"model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412}],"score":62.04,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":24700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1000},"family":"Humanity Sixth Sense","row_ids":["fe8e2c3470907e5adda03a06d7b5f363"],"score":37.04},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1200},"family":"MMMU-Pro","row_ids":["71c04ca8ce2b90585afee7d563834b34"],"score":70.59}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","included":true,"model_id":"m-db208dec31031420","original_version_label":"current public protocol","peer_count":28,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","included":true,"model_id":"m-db208dec31031420","original_version_label":"current Standard 10","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295}],"score":53.81,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","included":true,"model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":67.8,"percentile_exact":{"denominator":59,"numerator":4000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","included":true,"model_id":"m-db208dec31031420","original_version_label":"0.1","peer_count":53,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2.1","peer_count":82,"percentile":61.73,"percentile_exact":{"denominator":81,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","included":true,"model_id":"m-db208dec31031420","original_version_label":"4.0","peer_count":76,"percentile":54.67,"percentile_exact":{"denominator":3,"numerator":164},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","included":true,"model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":78,"percentile":58.44,"percentile_exact":{"denominator":77,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","included":true,"model_id":"m-db208dec31031420","original_version_label":"2.0","peer_count":50,"percentile":34.69,"percentile_exact":{"denominator":49,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","included":true,"model_id":"m-db208dec31031420","original_version_label":"1.1","peer_count":27,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","included":true,"model_id":"m-db208dec31031420","original_version_label":"current AA protocol","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","included":true,"model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","included":true,"model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","included":true,"model_id":"m-db208dec31031420","original_version_label":"current public protocol","peer_count":28,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","included":true,"model_id":"m-db208dec31031420","original_version_label":"current Standard 10","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295}],"result_updated_at":"2026-10-15","score":48.62,"score_exact":{"denominator":9172073422512,"numerator":445978209583391}}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"source_id":"meta/muse-spark-open"},"m-dea86e466cdc70f2":{"id":"m-dea86e466cdc70f2","in_window":true,"name":"Gemini 4 Carbon","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench 4.0 (AA)","benchmark_version":"4.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"01c0338617b2ef52cd797f5b878bbe71","model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench 4.0 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"6dfc6fa538421f2fee6d35c4","raw_value":65.2,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"65.2%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench 4.0 (AA):Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench:Terminal-Bench 4.0 (AA)","source_record_id":"carbon-image:Terminal-Bench 4.0 (AA):Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":65.2},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench Zapier","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"3bdbfd4c6e52ba4e0916b5c06403b6c3","model_id":"m-dea86e466cdc70f2","original_version_label":"current published developer protocol","profile_key":"native","protocol_description":"Official comparison image; AutomationBench Zapier; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"f658b6663efcee3dc42343bd","raw_value":53.9,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"53.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:AutomationBench Zapier:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:AutomationBench:AutomationBench Zapier","source_record_id":"carbon-image:AutomationBench Zapier:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":53.9},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science 0.1 (AA)","benchmark_version":"0.1.0","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"reasoning","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"77b00ca561419071138d999c141114f3","model_id":"m-dea86e466cdc70f2","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Official comparison image; Terminal-Bench-Science 0.1 (AA); original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"4f191cd5eb9caf57e48bc0e7","raw_value":64.8,"research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"64.8%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Terminal-Bench-Science 0.1 (AA):Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Terminal-Bench-Science:Terminal-Bench-Science 0.1 (AA)","source_record_id":"carbon-image:Terminal-Bench-Science 0.1 (AA):Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":64.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6a9666909d8f24f5a42848b17fcf718","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1601.74,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1601.74},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62d20dc1da6731366e64e80ccede8f11","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1567.14,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1567.14},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"285964e342882d8af4d5723bf64de4f6","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"df2bb91da1d8dcf5a57c997934d3d0df","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49233333333333335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1d6f53f1c5ceb7bda7e240fe7aefda31","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9575718015665796,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9575718015665796},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3982eacf6e1edec7d7467ca714480a5e","model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8023637557452397,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8023637557452397},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4388b84065fde96ec6cb1e998bd1706c","model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7759585317030244,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7759585317030244},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"55eb699906b0c7c1b3b10d4a09111d45","model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6345940610087639,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6345940610087639},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d9d5f29d99109876a8b77faf6dce557","model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.623809523809524,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.623809523809524},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1f9d71f5a674a42829e5f163592cd1eb","model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.619047619047619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.619047619047619},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2b138ff30d3aea61538d499ad3780a74","model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.394523232323232,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.394523232323232},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b6023def22e5f4a16cedd62ea8d56d0","model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.296189898989899,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296189898989899},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9f9874e0bf174fbb15ae77235e9dbcfe","model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.674,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.674},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"be9200177da44b3d4145725159c51eeb","model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.619,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.619},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"51ef194d229111ae4e3dbc5ebd2dca38","model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.296,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d56e2d0176f86933d4bb8463527005e2","model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dd386b581ae0af143558c88f7ca9719","model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.606060606060606,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.606060606060606},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8dde2603c050365ba8458b5b95c47df3","model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.651515151515151,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-carbon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.651515151515151},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f92157f7e16758df213ac054478d5675","model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.626262626262626,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.626262626262626},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1d92d6e42453c0088d0102f38e9fbca","model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"91ffcce17887e30ef05c459506897ec3","model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.65,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.65},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":153,"numerator":15100},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench-Science","row_ids":["416c079f92e7711d76f35d3a6da3f408"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"HLE","row_ids":["2541b171f3a9b70175c5eff4ce5a2a47"],"score":100.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4900},"family":"CritPt","row_ids":["29eed13fef0ca4e5917825122c33b6b0"],"score":96.08}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","peer_count":64,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","peer_count":52,"percentile":96.08,"percentile_exact":{"denominator":51,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69}],"score":98.69,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":3564,"numerator":321485},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7400},"family":"GDPval-AA","row_ids":["5c7968ea3f6404bb3d978285ac42aa05"],"score":91.36},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":375},"family":"Terminal-Bench","row_ids":["2db4ed4fb97d84958b935e83df5db43d"],"score":93.75},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["1b1f1f3e062850814bcd282684d4e879"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":1000},"family":"AA-Briefcase","row_ids":["fe408037e14234a002d5a2d7a3b186f0"],"score":90.91},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":75},"family":"OSWorld","row_ids":["f7b8bce78b6bbb414e44451ecece6456"],"score":75.0}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","peer_count":82,"percentile":91.36,"percentile_exact":{"denominator":81,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","peer_count":49,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":78,"percentile":90.91,"percentile_exact":{"denominator":11,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5}],"score":90.2,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":4400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4400},"family":"FrontierCode","row_ids":["54642e364cbdcff99c9cf0483b73a212"],"score":89.8}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","peer_count":50,"percentile":89.8,"percentile_exact":{"denominator":49,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061}],"score":89.8,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":94350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1750},"family":"GDP.pdf","row_ids":["2ef7660125e00e7ecf9ebbabeb38c644"],"score":92.11},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":5050},"family":"AA-Omniscience","row_ids":["00d59d147ff0b7cd3c12bdcaa9a4b516","91da142d2c0422108afa80f33bfb9bb4"],"score":95.28}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339}],"score":93.69,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":100},"family":"MMMU-Pro","row_ids":["34112c44d835a7ef36f154c3c42aa2b1"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["98879b4fa09dc7b005731381bea066e0"],"score":100.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5}],"score":100.0,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","peer_count":64,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","peer_count":52,"percentile":96.08,"percentile_exact":{"denominator":51,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","peer_count":82,"percentile":91.36,"percentile_exact":{"denominator":81,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","peer_count":49,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":78,"percentile":90.91,"percentile_exact":{"denominator":11,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","peer_count":50,"percentile":89.8,"percentile_exact":{"denominator":49,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5}],"result_updated_at":"2026-10-28","score":93.92,"score_exact":{"denominator":11958374736,"numerator":1123155312275}}],"release_date":"2026-10-28","release_status":"published","release_verified":true,"source_id":"published/aa:gemini-4-carbon"},"m-e060cf24e4176269":{"id":"m-e060cf24e4176269","in_window":true,"name":"Agnes 2.5 Pro Beta","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","model_id":"m-e060cf24e4176269","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":4,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":41.67,"evaluator_count":1,"exact":{"denominator":59,"numerator":4750},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4750},"family":"GPQA Diamond","row_ids":["2f5e3b6ebf2112f4f734c9c632e248bf"],"score":80.51}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505}],"score":80.51,"weight":25},{"effective_weight_percent":41.67,"evaluator_count":1,"exact":{"denominator":54,"numerator":3625},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":75},"family":"tau3-Banking","row_ids":["569f96b9f64d6f89c3e721fb138a9646"],"score":75.0},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1600},"family":"GDPval-AA","row_ids":["67407c66c8ccc8ec87ccd4b6ae121aa3"],"score":59.26}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"v2.1","peer_count":82,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33}],"score":67.13,"weight":25},{"effective_weight_percent":16.67,"evaluator_count":1,"exact":{"denominator":17,"numerator":550},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":550},"family":"MMMU-Pro","row_ids":["d7ea5990ff16dd400cf1a637a1392f4c"],"score":32.35}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607}],"score":32.35,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"v2.1","peer_count":82,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607}],"result_updated_at":null,"score":66.91,"score_exact":{"denominator":649944,"numerator":43486475}}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"source_id":"published/aa:agnes-2-5-pro-beta"},"m-e1cba225d1430824":{"id":"m-e1cba225d1430824","in_window":true,"name":"Qwen3.8-Flash-Next","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","model_id":"m-e1cba225d1430824","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","model_id":"m-e1cba225d1430824","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","model_id":"m-e1cba225d1430824","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","model_id":"m-e1cba225d1430824","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","model_id":"m-e1cba225d1430824","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","model_id":"m-e1cba225d1430824","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","model_id":"m-e1cba225d1430824","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3717,"numerator":264100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5200},"family":"GPQA Diamond","row_ids":["c6f5e59b8c457f8fd73cf7fe23e60786"],"score":88.14},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3400},"family":"HLE","row_ids":["1e856649fb09c4da785f8c4ea8ad05e0"],"score":53.97}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":88.14,"percentile_exact":{"denominator":59,"numerator":5200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond","peer_count":64,"percentile":53.97,"percentile_exact":{"denominator":63,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163}],"score":71.05,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":603860552},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1150},"family":"tau3-Banking","row_ids":["24d253e4fe3d6de3cc614f0a892166b1"],"score":88.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6700},"family":"GDPval-AA","row_ids":["c47c79a0ac9d03dc0858a3e32eb10cf4"],"score":82.72},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":68},"family":"Terminal-Bench","row_ids":["0978fbc73a68e2ee70d02c130e0d1cff"],"score":68.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1250},"family":"AutomationBench","row_ids":["12fa4c01c24c67c3a86dd523987aed83"],"score":65.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6700},"family":"AA-Briefcase","row_ids":["62ba5fa448ec9e7825ee5093f17e487d"],"score":87.01}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v2.1","peer_count":82,"percentile":82.72,"percentile_exact":{"denominator":81,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"4.0","peer_count":76,"percentile":68.0,"percentile_exact":{"denominator":1,"numerator":68},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":78,"percentile":87.01,"percentile_exact":{"denominator":77,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51}],"score":78.4,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["80a064d01bdbcc1f636799e9dd23518a"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"1.1","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7}],"score":100.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":6042,"numerator":401725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["1bcc725977832eb3960634353511e9be"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2525},"family":"GDP.pdf","row_ids":["046ca38af4c5ad3c8241882797259027"],"score":66.45},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1750},"family":"AA-Omniscience","row_ids":["0351ad1258b35881ec0b312bb72c8d29","8283a5fc47f30f74b98c14b10fde0469"],"score":33.02}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129}],"score":66.49,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1050},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1050},"family":"MMMU-Pro","row_ids":["42c55648590ad4a69cdc2a21de6faac7"],"score":61.76}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676}],"score":61.76,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":88.14,"percentile_exact":{"denominator":59,"numerator":5200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond","peer_count":64,"percentile":53.97,"percentile_exact":{"denominator":63,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v2.1","peer_count":82,"percentile":82.72,"percentile_exact":{"denominator":81,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"4.0","peer_count":76,"percentile":68.0,"percentile_exact":{"denominator":1,"numerator":68},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":78,"percentile":87.01,"percentile_exact":{"denominator":77,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"1.1","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676}],"result_updated_at":"2026-08-26","score":76.84,"score_exact":{"denominator":818935127010,"numerator":62923865665709}}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.8-flash-next"},"m-e5d076977e951607":{"id":"m-e5d076977e951607","in_window":true,"name":"GPT-6 Luna","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","model_id":"m-e5d076977e951607","original_version_label":"1.1","profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"c37eddc526d0ea268561864cec601165","model_id":"m-e5d076977e951607","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1437.0,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"1437","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row2:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-cdb1d51fae04:table0:row2:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"分","value":1437.0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"be25bcbd47f8306b2da2234ede387aa9","model_id":"m-e5d076977e951607","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Knowledge work AA-Briefcase v1.1; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1336.0,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"1336","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row4:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-cdb1d51fae04:table0:row4:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":1336.0},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","model_id":"m-e5d076977e951607","original_version_label":"2.1","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"c6b0cf71a2d3dc34173c226f0e2d751d","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Agentic coding Terminal-Bench 4.0; preserve page footnotes and task-specific settings","protocol_id":"25f728ad48355959443be093","raw_value":16.4,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"16.4%","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row11:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Terminal-Bench:Agentic coding Terminal-Bench 4.0","source_record_id":"official-cdb1d51fae04:table0:row11:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":16.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","model_id":"m-e5d076977e951607","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a67f5a0d13a3ef07b45639befa77b60c","model_id":"m-e5d076977e951607","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row27","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","model_id":"m-e5d076977e951607","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f1e82ec9268430b74f667de2ed65dceb","model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797109826589595,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna-xhigh:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797109826589595},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"0281315421da8621fd6deffc4ab4200b","model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.778612716763006,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna-medium:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.778612716763006},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fbaf699d4f9948fbf8d971ba8cbf893b","model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.620809248554913,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.620809248554913},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"a34e0f06622e7287dda0220cb718389c","model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.77514450867052,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna-high:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77514450867052},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"850650c0d70d10d4a701c83c126e3cfc","model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.759537572254335,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.759537572254335},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77c765575b9f84b169af4c22abe2cef2","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1235.49,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna-xhigh:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1235.49},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"529cb62ec57d64b5ee17455aef69bb3f","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1096.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1096.09},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6db5032f068e2f31ca40aaa300788377","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":903.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna-non-reasoning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":903.05},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b685ad52a2508c9cb36efd9844a360f","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1186.6,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna-high:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1186.6},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"012b084803a2b44d3d485cb1cdc28219","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":740.21,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna-low:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":740.21},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"80f91ad3a93a4ba9aeafa40b4b004de9","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.386,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-xhigh:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.386},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"296fd364581c75e65318dbb44393590c","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.387,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.387},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9e86cc0bcb40148727bec04837f9cc5b","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.301,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-non-reasoning:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.301},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7150c1e8fb085adae42357f4a7b146ad","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-high:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4c547408adb3c811a5b372cb58cd7a6","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.361,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-low:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.361},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5e6bd0432ec3aec25c7708c68f850c32","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1601520086862106,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-xhigh:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1601520086862106},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a13a09d28ccbab420878f778cab84e27","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14736269711799888,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14736269711799888},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cccc66ece2e40ceb1affce2a13ec27fd","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20028612303290416,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-non-reasoning:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20028612303290416},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c3f8e16a65cfb327ce1907c3a0922bc","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.15081967213114755,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-high:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15081967213114755},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"52d7a08fbee352d3e53e0e6df2f83c3a","model_id":"m-e5d076977e951607","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1517996870109546,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna-low:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1517996870109546},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dd085db97f0d70a411936e0dcda7727d","model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.47800686767566575,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna-xhigh:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47800686767566575},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8126a0ab65d3182534c4e483a68946e","model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4046136879135791,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4046136879135791},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f47ac0aa86663de94286c15c427d10b7","model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.09141771078680867,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna-non-reasoning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09141771078680867},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7be091719301381b62844b65e09526d0","model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4782449139916615,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna-high:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4782449139916615},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0eaef8368ed38947020e1dc63c18b","model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.12049187272690505,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna-low:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12049187272690505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c85dabb546ec1c35892d98d787840de","model_id":"m-e5d076977e951607","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0523809523809524,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna-xhigh:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0523809523809524},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a8f402b1825ea6e2887430214bcd75bc","model_id":"m-e5d076977e951607","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0142857142857143,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0142857142857143},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8fc3c64575d5beee3ba36d781dc2c873","model_id":"m-e5d076977e951607","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna-non-reasoning:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","model_id":"m-e5d076977e951607","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"68c36ea1426e59149731b66828b675ad","model_id":"m-e5d076977e951607","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0285714285714286,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna-high:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0285714285714286},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c10c1f9236de38227d8b12149bf3370","model_id":"m-e5d076977e951607","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna-low:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e6459b54cde5cc3a522aedf20a70d8d6","model_id":"m-e5d076977e951607","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.184238888888889,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna-xhigh:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.184238888888889},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b30b21d7ef196e2e60bd20c40d898c2b","model_id":"m-e5d076977e951607","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.127693181818182,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.127693181818182},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"809a248c9a8211973725062f59cc09b5","model_id":"m-e5d076977e951607","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0158234848484848,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna-non-reasoning:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0158234848484848},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","model_id":"m-e5d076977e951607","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ff058b64f2c6836b1b7a7720cdb5b87e","model_id":"m-e5d076977e951607","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.186015151515152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna-high:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186015151515152},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9271a716a1ad6e1eb09c76ec51e0313e","model_id":"m-e5d076977e951607","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0219545454545455,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna-low:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0219545454545455},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d05751f753832e6948a433cb24e30d48","model_id":"m-e5d076977e951607","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.189,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna-xhigh:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.189},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"24c26a94e6c645990758badfc57e4ef3","model_id":"m-e5d076977e951607","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.151,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.151},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"684a339ee9c3ad66cc983f87b4b3a537","model_id":"m-e5d076977e951607","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.063,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna-non-reasoning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.063},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","model_id":"m-e5d076977e951607","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3bb4231ab5c1518404d27fc034f89c63","model_id":"m-e5d076977e951607","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.156,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna-high:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5ae62065a38417e3fdceaea7742e9106","model_id":"m-e5d076977e951607","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna-low:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"59d9aa00ce93f703ea4e16e0fe1f14cc","model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna-xhigh:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"31c3a40e91efc47278344ebc1c17bda6","model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.138,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.138},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"04ecdb407074bd14821b286cdb5dae57","model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.058,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna-non-reasoning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.058},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5b4871ef41f9fc9ed31d5a830bf31d47","model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.182,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna-high:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a9b1cc589df047d38683a429a5c3ca34","model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna-low:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"25edc6016fa062e9b15580bdf8efc4e8","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-luna-xhigh:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a379c8ea4c01c18381c17cfe48b4b53","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0252525252525253,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-luna-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0252525252525253},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"03ca642609ebdfc5d25b13c2ea93c6fa","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-luna-non-reasoning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f54a9b9328eeb87160c6fecff9cdb2e5","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-luna:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a3c1f2c30f48feed5b02716dbdf40854","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0454545454545455,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-luna-high:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0454545454545455},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ddeb62e7fd3892e5bbdccc5ef24cb2cc","model_id":"m-e5d076977e951607","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-luna-low:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0c9ebd5535e7f16892a9ba18f96ae880","model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.37,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna-xhigh:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb3204c0b953048ba92db66a5713edab","model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.275,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a9bc9bd469b5eeec73ad9613b05d8b4","model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna-non-reasoning:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"31cb2813e245fe0112d5bf75a409923a","model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.31,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna-high:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63c247f23f68c882615bc47c30d36d6d","model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.095,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna-low:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.095},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49005d88bd0ac4d3a4c6facc3398820a","model_id":"m-e5d076977e951607","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1372.52,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-xhigh:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna-xhigh:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1372.52},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"271805586aacfdf5b4cd59714f53bf19","model_id":"m-e5d076977e951607","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1270.45,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1270.45},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf83e7f1bb6193d722ddd5255e08a890","model_id":"m-e5d076977e951607","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1051.67,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-non-reasoning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna-non-reasoning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1051.67},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","model_id":"m-e5d076977e951607","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"83383f94a465081d737504f6a663d2b6","model_id":"m-e5d076977e951607","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1352.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-high:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna-high:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1352.1},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b76426811bf6f5f24ad13e19c9319b8","model_id":"m-e5d076977e951607","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1040.48,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna-low:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna-low:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1040.48},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":663,"numerator":45650},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["6de4c10b838de9bb43446d62c1dedc6c"],"score":65.38},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":200},"family":"HLE","row_ids":["ca6b622f3e2324e23758731c3eedb79c"],"score":66.67},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3800},"family":"CritPt","row_ids":["a69315d97d8db2867eda5787e5741424"],"score":74.51}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"0.1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"Diamond","peer_count":64,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","peer_count":52,"percentile":74.51,"percentile_exact":{"denominator":51,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425}],"score":68.85,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":32706828,"numerator":1477535075},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5200},"family":"GDPval-AA","row_ids":["f4a874da20b414e345395af6876bfede"],"score":64.2},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":175},"family":"Terminal-Bench","row_ids":["a8bde16affe89631785f57ec46935865"],"score":29.17},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1225},"family":"AutomationBench","row_ids":["c468f61930c57074fdae0429fd64cde2"],"score":64.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4700},"family":"AA-Briefcase","row_ids":["27c5dd58443826f50a40adf5e5fbdf05"],"score":61.04},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"OSWorld","row_ids":["b4bc853a9a856396a7313b82125218ee"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1200},"family":"AA-AnalystAgent","row_ids":["ebd2172910fbd66a3c4a6655aa5bdef4"],"score":52.17}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2.1","peer_count":82,"percentile":64.2,"percentile_exact":{"denominator":81,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","included":true,"model_id":"m-e5d076977e951607","original_version_label":"4.0","peer_count":49,"percentile":29.17,"percentile_exact":{"denominator":6,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","included":true,"model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":78,"percentile":61.04,"percentile_exact":{"denominator":77,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.1","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","peer_count":24,"percentile":52.17,"percentile_exact":{"denominator":23,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375}],"score":45.18,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":4,"exact":{"denominator":1274,"numerator":32475},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3300},"family":"FrontierCode","row_ids":["91532a2952c890cc776404a1ca00195c"],"score":67.35},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":450},"family":"Vals Code Migration","row_ids":["1045763a4cc14f13d71cbc23b34d8b21"],"score":34.62},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["12d12a117fed600435582eb0cdccab38"],"score":0.0},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":0},"family":"MirrorCode","row_ids":["9b168b30e93ee9c993180224bf1ba997"],"score":0.0}],"family_count":4,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.0","peer_count":50,"percentile":67.35,"percentile_exact":{"denominator":49,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":27,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666}],"score":25.49,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":1755175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3225},"family":"GDP.pdf","row_ids":["ff1f819163bf03455ec9cbe6d10872e3"],"score":84.87},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":450},"family":"SimpleQA Verified","row_ids":["d183987f58d811d89f8c42a38fe26df4"],"score":26.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["efcb1058be3817d672ab8961f07089a0","7ec06e89d9178a8c5c27508ed35805db"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906}],"score":51.26,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":51,"numerator":1000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Humanity Sixth Sense","row_ids":["c23775cd50aa5f64341fa34e8b1cf6d3"],"score":0.0},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1000},"family":"MMMU-Pro","row_ids":["7443158d8b6f105fcb7949451afd84ee"],"score":58.82},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["4a70f456ec596dc9670786d2fcc37f1a"],"score":0.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current public protocol","peer_count":28,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","included":true,"model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1}],"score":19.61,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"0.1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"Diamond","peer_count":64,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","peer_count":52,"percentile":74.51,"percentile_exact":{"denominator":51,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2.1","peer_count":82,"percentile":64.2,"percentile_exact":{"denominator":81,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","included":true,"model_id":"m-e5d076977e951607","original_version_label":"4.0","peer_count":49,"percentile":29.17,"percentile_exact":{"denominator":6,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","included":true,"model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":78,"percentile":61.04,"percentile_exact":{"denominator":77,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.1","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","peer_count":24,"percentile":52.17,"percentile_exact":{"denominator":23,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.0","peer_count":50,"percentile":67.35,"percentile_exact":{"denominator":49,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":27,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current public protocol","peer_count":28,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","included":true,"model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1}],"result_updated_at":"2026-10-26","score":45.82,"score_exact":{"denominator":10726662138192,"numerator":491483807341325}}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"source_id":"openai/gpt-6-luna"},"m-e7f69d0dac9fe632":{"id":"m-e7f69d0dac9fe632","in_window":true,"name":"Gemini 3.5 Flash-Lite","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":365092,"numerator":10439225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3000},"family":"GPQA Diamond","row_ids":["f63fe95853d236ed53d2b79d02b859cb"],"score":50.85},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["2c2a68b20b30379e3063400c424fa6d0"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":200},"family":"HLE","row_ids":["cda0d514d2e746d7e415df4602a30e61"],"score":28.57},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":300},"family":"CritPt","row_ids":["1b14f3e78b8a7ccb2a6b6d0eda26f9e2"],"score":17.65}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":50.85,"percentile_exact":{"denominator":59,"numerator":3000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","peer_count":64,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","peer_count":52,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035}],"score":28.59,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1027026,"numerator":39546907},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["62cc0b9af0e4241c44026a470cdd621c"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":100},"family":"GDPval-AA","row_ids":["9d93ac6bf71dd0301098e7d1b1db2fb3"],"score":33.33},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["54cd2ea0dc02b5047f97de0db07066fb"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":875},"family":"AutomationBench","row_ids":["bdc1d1c2838fa6785d600f455b24f788"],"score":46.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1500},"family":"AA-Briefcase","row_ids":["0ec16dc40db0414f2891438008ea0a1a"],"score":19.48},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":500},"family":"EnterpriseOps-Gym","row_ids":["5ffca0a973b63389e8c0ae797941349a"],"score":55.56}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","peer_count":82,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":46.05,"percentile_exact":{"denominator":19,"numerator":875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":78,"percentile":19.48,"percentile_exact":{"denominator":77,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188}],"score":38.51,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":1500},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1500},"family":"FrontierCode","row_ids":["26e16bd219d1e573463da6c55739a16d"],"score":30.61}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","peer_count":50,"percentile":30.61,"percentile_exact":{"denominator":49,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323}],"score":30.61,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":119275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1175},"family":"GDP.pdf","row_ids":["b1d8307905906d9eadee1409ffb734e6"],"score":61.84},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3000},"family":"AA-Omniscience","row_ids":["5ac3b36a409b29ee98389b739f2f25b7","4347afed89affe307401c55cb1f477b8"],"score":56.6}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043}],"score":59.22,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["ef305cf290c211e130604e205ee97f46"],"score":50.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624}],"score":50.0,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":50.85,"percentile_exact":{"denominator":59,"numerator":3000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","peer_count":64,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","peer_count":52,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","peer_count":82,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":46.05,"percentile_exact":{"denominator":19,"numerator":875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":78,"percentile":19.48,"percentile_exact":{"denominator":77,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","peer_count":50,"percentile":30.61,"percentile_exact":{"denominator":49,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624}],"result_updated_at":null,"score":39.74,"score_exact":{"denominator":3057357807504,"numerator":121505356698007}}],"release_date":"2026-07-21","release_status":"published","release_verified":true,"source_id":"google/gemini-3.5-flash-lite"},"m-f33f6c88c27b2b9a":{"id":"m-f33f6c88c27b2b9a","in_window":true,"name":"Inkling 2 Small","observed_records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206}],"official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-11-05","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":5967,"numerator":413600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":800},"family":"Terminal-Bench-Science","row_ids":["7ead6883bce4c045cf26f5b116b4cd48"],"score":61.54},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":700},"family":"HLE","row_ids":["89d939473fcae282b052740fc3fcfd7a"],"score":77.78},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3500},"family":"CritPt","row_ids":["34dffe7c2a7d437da340dfb5246fe9c8"],"score":68.63}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","peer_count":64,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","peer_count":52,"percentile":68.63,"percentile_exact":{"denominator":51,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"score":69.31,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":35432397,"numerator":2557154263},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1050},"family":"tau3-Banking","row_ids":["66284fe19cdf80bb97a995a814e23753"],"score":80.77},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6400},"family":"GDPval-AA","row_ids":["f53e9f08adf2987f0e2c59d35bf57232"],"score":79.01},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":184},"family":"Terminal-Bench","row_ids":["b334bad67fbd4a96eed8e669e0111fbb"],"score":61.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1025},"family":"AutomationBench","row_ids":["fb9e5b876d81fca5e8cf53e8169709cb"],"score":53.95},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":900},"family":"AA-Briefcase","row_ids":["ae0bf141fcba5d508b0da173dcf01538"],"score":81.82},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":650},"family":"EnterpriseOps-Gym","row_ids":["ccd3047c248ced9697a972d3c394d08d"],"score":72.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1750},"family":"AA-AnalystAgent","row_ids":["e23e94bce569e3668ea49a1736f85e07"],"score":76.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","peer_count":82,"percentile":79.01,"percentile_exact":{"denominator":81,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","peer_count":76,"percentile":61.33,"percentile_exact":{"denominator":3,"numerator":184},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":78,"percentile":81.82,"percentile_exact":{"denominator":11,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125}],"score":72.17,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":3400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3400},"family":"FrontierCode","row_ids":["0fe380f78c53ebf4d268a83df481e0ad"],"score":69.39}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","peer_count":50,"percentile":69.39,"percentile_exact":{"denominator":49,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101}],"score":69.39,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":77450},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1650},"family":"GDP.pdf","row_ids":["792f8618c792ce48ee9c7c6155488bb6"],"score":86.84},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3550},"family":"AA-Omniscience","row_ids":["28a0e72d64e34dbdaef952ff1ff713b4","1a5d81b92d3a8437744af4b44b483171"],"score":66.98}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385}],"score":76.91,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1250},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1250},"family":"MMMU-Pro","row_ids":["e681be1aa8a9a82be96c894925b92cdd"],"score":73.53}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185}],"score":73.53,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","peer_count":64,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","peer_count":52,"percentile":68.63,"percentile_exact":{"denominator":51,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","peer_count":82,"percentile":79.01,"percentile_exact":{"denominator":81,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","peer_count":76,"percentile":61.33,"percentile_exact":{"denominator":3,"numerator":184},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":78,"percentile":81.82,"percentile_exact":{"denominator":11,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","peer_count":50,"percentile":69.39,"percentile_exact":{"denominator":49,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185}],"result_updated_at":null,"score":71.98,"score_exact":{"denominator":893888511516,"numerator":64345620454681}}],"release_date":"2026-11-05","release_status":"published","release_verified":true,"source_id":"published/aa:inkling-small-1102-rc17"},"m-f74476b1532b3a4f":{"id":"m-f74476b1532b3a4f","in_window":true,"name":"Ling 3.1 Flash","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":5,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":26334,"numerator":1972813},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":700},"family":"GDPval-AA","row_ids":["efd709c75b89ba20347f69709bb9a3a4"],"score":77.78},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":78},"family":"Terminal-Bench","row_ids":["070c8b1a9e5fb50b3ddeb6aba50faf78"],"score":78.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1500},"family":"AutomationBench","row_ids":["712bffe9ecc8daeca7e0dc7a1fe07e58"],"score":78.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5000},"family":"AA-Briefcase","row_ids":["d6fbe929d324cc028c3c012f923590ba"],"score":64.94}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","peer_count":82,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","peer_count":76,"percentile":78.0,"percentile_exact":{"denominator":1,"numerator":78},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","peer_count":78,"percentile":64.94,"percentile_exact":{"denominator":77,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94}],"score":74.92,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":38,"numerator":1625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1625},"family":"GDP.pdf","row_ids":["d7e50c008783e8a4b24ebddf0211490f"],"score":42.76}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"score":42.76,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","peer_count":82,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","peer_count":76,"percentile":78.0,"percentile_exact":{"denominator":1,"numerator":78},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","peer_count":78,"percentile":64.94,"percentile_exact":{"denominator":77,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"result_updated_at":null,"score":60.63,"score_exact":{"denominator":237006,"numerator":14368565}}],"release_date":"2026-10-01","release_status":"published","release_verified":true,"source_id":"inclusionai/ling-3.1-flash"},"m-f892623396612d9b":{"id":"m-f892623396612d9b","in_window":true,"name":"Step 5 Preview","observed_records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","model_id":"m-f892623396612d9b","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","model_id":"m-f892623396612d9b","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","model_id":"m-f892623396612d9b","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","model_id":"m-f892623396612d9b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","model_id":"m-f892623396612d9b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","model_id":"m-f892623396612d9b","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2b8cf679ed4d4658129609f163b9f607","model_id":"m-f892623396612d9b","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0238095238095238,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"step-5:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0238095238095238},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","model_id":"m-f892623396612d9b","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","model_id":"m-f892623396612d9b","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c3c1a9e38c3fc762089faf1cd02d09ab","model_id":"m-f892623396612d9b","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"step-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","model_id":"m-f892623396612d9b","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","model_id":"m-f892623396612d9b","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","model_id":"m-f892623396612d9b","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"official_evidence":{"artifact_sha256":"f316d16c93b2d4ce98e733917634947e32a9218037b9e1d3d427de4b75275ddb","basis":"official_release_page","date":"2026-09-20","sha256":"f316d16c93b2d4ce98e733917634947e32a9218037b9e1d3d427de4b75275ddb","url":"https://www.stepfun.com/news/step-5-preview"},"primary_profile_id":"native","profiles":[{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":20349,"numerator":1015100},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":350},"family":"Terminal-Bench-Science","row_ids":["f80ea7a9c8b1ad0f866390a070dec6ae"],"score":18.42},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":500},"family":"HLE","row_ids":["ea28428a53ddb22639d7f5b7e50547a2"],"score":71.43},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3050},"family":"CritPt","row_ids":["b5e83ad92c83f666dbcd3cdc392d63b6"],"score":59.8}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","included":true,"model_id":"m-f892623396612d9b","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","included":true,"model_id":"m-f892623396612d9b","original_version_label":"Diamond","peer_count":64,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325}],"score":49.88,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":152631864,"numerator":9897530725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"APEX-Agents","row_ids":["519c33c19b165c6197391af117e522e3"],"score":75.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5900},"family":"GDPval-AA","row_ids":["d52704f212d466a4d653d53f1794d1db"],"score":72.84},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1375},"family":"Terminal-Bench","row_ids":["d5a9a67230f3e72676fe2415a004dd47"],"score":57.29},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1175},"family":"AutomationBench","row_ids":["85e95115c24e6ebd5d7a72360c176406"],"score":61.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5300},"family":"AA-Briefcase","row_ids":["5342274720726a814f5f5d822f07b510"],"score":68.83},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":250},"family":"EnterpriseOps-Gym","row_ids":["6be4e7acd5ffa2c31291fe47e51a2a60"],"score":83.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":800},"family":"AA-AnalystAgent","row_ids":["1f372ec6dfb68a53b6ae42bbb816dddf"],"score":34.78}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2.1","peer_count":82,"percentile":72.84,"percentile_exact":{"denominator":81,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","included":true,"model_id":"m-f892623396612d9b","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","included":true,"model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":78,"percentile":68.83,"percentile_exact":{"denominator":77,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","peer_count":24,"percentile":34.78,"percentile_exact":{"denominator":23,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"score":64.85,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":25250},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2000},"family":"FrontierCode","row_ids":["4086551829a96ee8ea164e44d1e7c1c5"],"score":40.82},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":500},"family":"Vals Code Migration","row_ids":["68fce01452fe549878e8933d04edaa05"],"score":38.46}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","included":true,"model_id":"m-f892623396612d9b","original_version_label":"2.0","peer_count":50,"percentile":40.82,"percentile_exact":{"denominator":49,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","included":true,"model_id":"m-f892623396612d9b","original_version_label":"1.1","peer_count":27,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0}],"score":39.64,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":62200},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1200},"family":"GDP.pdf","row_ids":["32f9ed0b80b3e83e515a9f3fb34bc9b3"],"score":63.16},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3200},"family":"AA-Omniscience","row_ids":["ec9021441721b216b9626f97b96062f1","880d1169acb398e66460aed2eb721de0"],"score":60.38}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072}],"score":61.77,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":650},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":650},"family":"MMMU-Pro","row_ids":["e0d9cab1ce675bebaf2c47a0db9ddea9"],"score":38.24}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902}],"score":38.24,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","included":true,"model_id":"m-f892623396612d9b","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","included":true,"model_id":"m-f892623396612d9b","original_version_label":"Diamond","peer_count":64,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2.1","peer_count":82,"percentile":72.84,"percentile_exact":{"denominator":81,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","included":true,"model_id":"m-f892623396612d9b","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","included":true,"model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":78,"percentile":68.83,"percentile_exact":{"denominator":77,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","peer_count":24,"percentile":34.78,"percentile_exact":{"denominator":23,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","included":true,"model_id":"m-f892623396612d9b","original_version_label":"2.0","peer_count":50,"percentile":40.82,"percentile_exact":{"denominator":49,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","included":true,"model_id":"m-f892623396612d9b","original_version_label":"1.1","peer_count":27,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902}],"result_updated_at":null,"score":52.79,"score_exact":{"denominator":376374110112,"numerator":19867815404455}}],"release_date":"2026-09-20","release_status":"published","release_verified":true,"source_id":"stepfun/step-5-preview"},"m-f8dbf8d16fd37aac":{"id":"m-f8dbf8d16fd37aac","in_window":true,"name":"GPT-5.6 Terra","observed_records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"707e27b848a443e1f331abbd2d46a87f","model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":80.7,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":80.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"dde4f14f270dc23fa94482cb2b80d4e8","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":68.3,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":68.3},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"692a7ec419b3a00c120affaf39ba525a","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3414141414141414,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4a0ee8f563e3a5cd3821037b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4a0ee8f563e3a5cd3821037b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3414141414141414},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"cbd8f25646bfd32d3fe9066e2d8d8b69","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2202020202020202,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6b010c46a952106e8addbd3"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"f6b010c46a952106e8addbd3","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2202020202020202},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"e52cf2658ea61d97b875e9cadefaf071","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.29797979797979796,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5b74da2db0d257625d093316"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"5b74da2db0d257625d093316","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.29797979797979796},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"fc712f1dbc71e4848fa5c2b7ef651a6d","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.35353535353535354,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa94921e720c05e0ef881ab4"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"fa94921e720c05e0ef881ab4","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.35353535353535354},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0}],"official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profiles":[{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":5525},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":50},"family":"FrontierMath","row_ids":["34d11ef45eddcdc63e92f205eb9eaa3c"],"score":50.0},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":75},"family":"GPQA Diamond","row_ids":["6b87188fa1e95a23261b2a5e267b3413"],"score":37.5},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["9d8fe865edc8da70092f0b94926485a3"],"score":57.89}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0}],"score":48.46,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":12,"numerator":275},"families":[{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":175},"family":"Terminal-Bench","row_ids":["7fbb0050443eb76fe54b1727ad2dc1fe"],"score":43.75},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"AutomationBench","row_ids":["e4ce479712f2cad43715215ba9603696"],"score":25.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"AA-Briefcase","row_ids":["56c8ab5d498b1025e9684eb22709b6c7"],"score":0.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","peer_count":49,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334}],"score":22.92,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":75},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":75},"family":"DeepSWE","row_ids":["99e7d048583a82e4719e8e299872ac27"],"score":37.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6}],"score":37.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"GDP.pdf","row_ids":["2f6ca6bd54419e351a2ba1a4c64dedd7"],"score":50.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7}],"score":50.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["2d9aead127b28a667d7f4b6c145f9945"],"score":7.41},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["079f06724159c79071056bf02b730573"],"score":50.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","peer_count":28,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0}],"score":28.7,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","peer_count":49,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","peer_count":28,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0}],"result_updated_at":"2026-07-09","score":38.22,"score_exact":{"denominator":8208,"numerator":313675}}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"source_id":"openai/gpt-5.6-terra"},"m-f936349cbdb1def9":{"id":"m-f936349cbdb1def9","in_window":true,"name":"DeepSeek V4 Pro 0813","observed_records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"8c629283bcf934e5a97ae61bcef9fec0","model_id":"m-f936349cbdb1def9","original_version_label":"4.0","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Terminal-Bench 4.0 (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"d3512b6f8cf5dcb54718109d","raw_value":12.4,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"12.4","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row9:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Terminal-Bench:Terminal-Bench 4.0 (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row9:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":12.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","model_id":"m-f936349cbdb1def9","original_version_label":"1.1","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"b7e9e3548b08b017e44f0d19beea862d","model_id":"m-f936349cbdb1def9","original_version_label":"published developer protocol","profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":43.2,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"43.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":43.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"dc36fb77534eeb86099f253b7683ccaf","model_id":"m-f936349cbdb1def9","original_version_label":"Diamond; fixed 198-question subset","profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Pro. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":92.4,"research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"92.4","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":92.4},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f23f1e17caba14beed3e77ae9ea4b43","model_id":"m-f936349cbdb1def9","profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":91.7,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"91.7","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / DeepSeek V4 Pro 0813"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":91.7},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"cbcb87856c45df67217ec3b0da377d9c","model_id":"m-f936349cbdb1def9","profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":32.4,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / DeepSeek V4 Pro 0813"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"7aef04ad579d81beb4f5c8ec2832a80c","model_id":"m-f936349cbdb1def9","profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":30.4,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"30.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / DeepSeek V4 Pro 0813"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":30.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"981ac64200af40a5afa217a11e6600b2","model_id":"m-f936349cbdb1def9","profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":58.8,"response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"58.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / DeepSeek V4 Pro 0813"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":58.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","model_id":"m-f936349cbdb1def9","original_version_label":"4.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","model_id":"m-f936349cbdb1def9","original_version_label":"1.1","profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb8131444f4b9b30dcdf9ea24c2fdcf5","model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":927.88,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro-non-reasoning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":927.88},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"45dc0541c71975743a6ec69a4e1d1cab","model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.386,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro-non-reasoning:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.386},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"285a98eaf45548b1df7ffceb7b75e55c","model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06080347448425627,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro-non-reasoning:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06080347448425627},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"86f8dc8aa01b7c521d7b2eb833b330bd","model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3490865332353694,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro-non-reasoning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3490865332353694},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f1eefd5ab642da9d01de8eac565cc324","model_id":"m-f936349cbdb1def9","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9a4228692fbd7ac9ccdfa3fcf72064f","model_id":"m-f936349cbdb1def9","original_version_label":"0.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","model_id":"m-f936349cbdb1def9","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f826bdea869f7f477dac594e5c0acaaa","model_id":"m-f936349cbdb1def9","original_version_label":"2.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.103058585858586,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.103058585858586},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e7362ebc8218df1b7c1a6a227f89086","model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.072,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro-non-reasoning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.072},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e8acc0fb3a761b4196c27a08de6e814","model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro-non-reasoning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f209459e33bb2ec173abeb800f71be3","model_id":"m-f936349cbdb1def9","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.141414141414141,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.141414141414141},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12ef68a3d464a7e89561e62fe4565302","model_id":"m-f936349cbdb1def9","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-pro-non-reasoning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"38b0678c0520f769448b83eb8adbd0cb","model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.085,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Non-reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62487f5522ba918d7a74813cceba9907","model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1066.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro-non-reasoning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro-non-reasoning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1066.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-13","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":17,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":800394,"numerator":45089825},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5350},"family":"GPQA Diamond","row_ids":["a4b58fb51111dbcc6b81f75ff3f7fc61"],"score":90.68},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["0898509511ccb710480bf98d3742bbb0"],"score":31.58},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1300},"family":"HLE","row_ids":["d543f8764a78abfabdc05d4f312b8b2f"],"score":61.9},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":700},"family":"CritPt","row_ids":["a2d7db240ac12326c1a9bfd85ff9b349"],"score":41.18}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","peer_count":64,"percentile":61.9,"percentile_exact":{"denominator":21,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","peer_count":52,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235}],"score":56.33,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":145236,"numerator":8347475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["76dde74584cddc5a2d8618786061726c"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1025},"family":"tau3-Banking","row_ids":["20bfc5477d34134d5c7c99539fd481a0"],"score":78.85},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"GDPval-AA","row_ids":["727c30106a7ec7d9e1c1205d514e6937"],"score":66.67},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":125},"family":"Terminal-Bench","row_ids":["35687651705594abdbf62890a0bd5a87"],"score":31.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1300},"family":"AutomationBench","row_ids":["63186f8028fad18deb2fa4c75b5a0888"],"score":68.42},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":400},"family":"AA-Briefcase","row_ids":["40fe84ac7f2d6389219a5c06f59da083"],"score":57.14},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["6c40f85770ab1fe03b45de2938f913b3"],"score":100.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":78.85,"percentile_exact":{"denominator":13,"numerator":1025},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","peer_count":82,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"4.0","peer_count":49,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":78,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731}],"score":57.48,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":441,"numerator":17000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2400},"family":"FrontierCode","row_ids":["2e337d0b2b65e72e819796e03cdc53e7"],"score":48.98},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"Vals Code Migration","row_ids":["7e542189106406c0f4ba769f7753c91e"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":50},"family":"DeepSWE","row_ids":["63840894bc67cdd6c5a9cd6e26e8b9d7"],"score":16.67}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"2.0","peer_count":50,"percentile":48.98,"percentile_exact":{"denominator":49,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":27,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7}],"score":38.55,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2485600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1000},"family":"GDP.pdf","row_ids":["c287cda1f48a055602b23132b6d39858"],"score":52.63},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":900},"family":"SimpleQA Verified","row_ids":["665cc51403349d6e59eed00558808309"],"score":52.94},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["3f8c786b8fc874f5d0f0ae1174917198","775332c591df3bb4219419f088bb5ff5"],"score":39.62}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429}],"score":48.4,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","peer_count":64,"percentile":61.9,"percentile_exact":{"denominator":21,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","peer_count":52,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":78.85,"percentile_exact":{"denominator":13,"numerator":1025},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","peer_count":82,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"4.0","peer_count":49,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":78,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"2.0","peer_count":50,"percentile":48.98,"percentile_exact":{"denominator":49,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":27,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429}],"result_updated_at":"2026-08-27","score":50.94,"score_exact":{"denominator":416912428296,"numerator":21235615018525}}],"release_date":"2026-08-13","release_status":"published","release_verified":true,"source_id":"deepseek/deepseek-v4-pro-0813"},"m-f99cd5ceed3a5b41":{"id":"m-f99cd5ceed3a5b41","in_window":true,"name":"Solar Mini 4","observed_records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8}],"official_evidence":{"artifact_sha256":"967417099d792a2cbc0abac952f572bc529067f8ba0587e3c041a5f253bc2a0e","basis":"official_release_page","date":"2026-10-01","sha256":"967417099d792a2cbc0abac952f572bc529067f8ba0587e3c041a5f253bc2a0e","url":"https://www.upstage.ai/blog/en/solar-mini-4"},"primary_profile_id":"native","profiles":[{"benchmark_count":6,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":7182,"numerator":275461},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1100},"family":"GDPval-AA","row_ids":["dde8616d382b2b0a447eb1104181da5e"],"score":40.74},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["7620c0fb57936004d0ff80edb6675f9f"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":800},"family":"AutomationBench","row_ids":["b1f8ac835c2c56f569ae7093d49e508f"],"score":42.11},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":200},"family":"AA-Briefcase","row_ids":["b76daa2b8d1a11ba3f7dd4198d33e0a0"],"score":28.57}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","peer_count":82,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":42.11,"percentile_exact":{"denominator":19,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":78,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74}],"score":38.35,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":4028,"numerator":113625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":925},"family":"GDP.pdf","row_ids":["d7e2c6c9bb5b766f6065fc5409408479"],"score":24.34},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1700},"family":"AA-Omniscience","row_ids":["8e92630a337eaf823218ceb658e6daf8","973d182535a01b720897f5a733a8d1b0"],"score":32.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599}],"score":28.21,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","peer_count":82,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":42.11,"percentile_exact":{"denominator":19,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":78,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599}],"result_updated_at":null,"score":33.85,"score_exact":{"denominator":3425814,"numerator":115947415}}],"release_date":"2026-10-01","release_status":"published","release_verified":true,"source_id":"upstage/solar-mini4"},"m-fb920161af1862b7":{"id":"m-fb920161af1862b7","in_window":true,"name":"Qwen3.8 2.4T A95B","observed_records":[{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","model_id":"m-fb920161af1862b7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","model_id":"m-fb920161af1862b7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","model_id":"m-fb920161af1862b7","original_version_label":"v1.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","model_id":"m-fb920161af1862b7","original_version_label":"Diamond","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","model_id":"m-fb920161af1862b7","original_version_label":"4.0","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","model_id":"m-fb920161af1862b7","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","model_id":"m-fb920161af1862b7","original_version_label":"current AA run","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":294400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5600},"family":"GPQA Diamond","row_ids":["b2e2268e4647b383f382e5a7860dafa3"],"score":94.92},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4000},"family":"HLE","row_ids":["b6295a8854af11f8a1045539e4932330"],"score":63.49}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":94.92,"percentile_exact":{"denominator":59,"numerator":5600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond","peer_count":64,"percentile":63.49,"percentile_exact":{"denominator":63,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201}],"score":79.2,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":10783773,"numerator":844623034},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1275},"family":"tau3-Banking","row_ids":["960dd18162897dbffed3123d8b9728ee"],"score":98.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6200},"family":"GDPval-AA","row_ids":["fcdd405fcadd197dbec16caec46c9124"],"score":76.54},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":56},"family":"Terminal-Bench","row_ids":["8981298b68f7a13ad078d01ac54f2b30"],"score":56.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1325},"family":"AutomationBench","row_ids":["f2b687ab9797dca77a0de360cefeb5cf"],"score":69.74},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5400},"family":"AA-Briefcase","row_ids":["be21483d79b65c8a6ad01a6eee1e0103"],"score":70.13},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"EnterpriseOps-Gym","row_ids":["32758e4ec324e0ec55fbc6f952cf8937"],"score":88.89},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":800},"family":"MCP Atlas","row_ids":["7b2615183b2725bb4693774f30af4a5b"],"score":88.89}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":98.08,"percentile_exact":{"denominator":13,"numerator":1275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v2.1","peer_count":82,"percentile":76.54,"percentile_exact":{"denominator":81,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"4.0","peer_count":76,"percentile":56.0,"percentile_exact":{"denominator":1,"numerator":56},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":78,"percentile":70.13,"percentile_exact":{"denominator":77,"numerator":5400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA run","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5}],"score":78.32,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":238325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2525},"family":"GDP.pdf","row_ids":["93e5cef565694d0b6ece46832b2fc2e8"],"score":66.45},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2750},"family":"AA-Omniscience","row_ids":["872cbb8006efe075a4c9b62b4f118011","7f66b653a79b36bb8c9fbcd69744e5f5"],"score":51.89}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656}],"score":59.17,"weight":20}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":94.92,"percentile_exact":{"denominator":59,"numerator":5600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond","peer_count":64,"percentile":63.49,"percentile_exact":{"denominator":63,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":98.08,"percentile_exact":{"denominator":13,"numerator":1275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v2.1","peer_count":82,"percentile":76.54,"percentile_exact":{"denominator":81,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"4.0","peer_count":76,"percentile":56.0,"percentile_exact":{"denominator":1,"numerator":56},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":78,"percentile":70.13,"percentile_exact":{"denominator":77,"numerator":5400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA run","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656}],"result_updated_at":null,"score":73.16,"score_exact":{"denominator":472092014394,"numerator":34540416783815}}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"source_id":"alibaba/qwen3.8-2.4t-a95b"},"m-fe0a26f770e1e011":{"id":"m-fe0a26f770e1e011","in_window":true,"name":"DiffusionGemma 26B A4B","observed_records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869}],"official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profiles":[{"benchmark_count":4,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":1000},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1000},"family":"GPQA Diamond","row_ids":["9804205b8a1d620728568c184381a38c"],"score":16.95}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":16.95,"percentile_exact":{"denominator":59,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869}],"score":16.95,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":27,"numerator":200},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":200},"family":"GDPval-AA","row_ids":["4fa0d328e940e0ba1d47d7e70c6e0521"],"score":7.41}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","peer_count":82,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5}],"score":7.41,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":175},"family":"IFBench","row_ids":["8e2367487c04d7937c7f24678241f073"],"score":43.75}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252}],"score":43.75,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":200},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":200},"family":"MMMU-Pro","row_ids":["d47a048ab5b9e3796a7da71c7471a272"],"score":11.76}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144}],"score":11.76,"weight":10}],"eligible":true,"profile_id":"native","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":16.95,"percentile_exact":{"denominator":59,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","peer_count":82,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144}],"result_updated_at":null,"score":20.02,"score_exact":{"denominator":433296,"numerator":8674375}}],"release_date":"2026-06-10","release_status":"published","release_verified":true,"source_id":"published/aa:diffusiongemma-26b-a4b"}},"eligible_observation_count":1536,"evidence_revision":"c15bb3a91f3f244b75b5e16fd7298c9ba896fd3779a94348148c7b36a8768542","method":{"basis":"Already published measurements from reviewed evaluators and official model release comparisons","configuration":"Task-specific published protocols; distinct precision profiles; no selection by score and no cross-setting maxima","dimensions":[{"key":"reasoning","label":"推理求解","weight":25},{"key":"agents","label":"工具与任务执行","weight":25},{"key":"coding","label":"编程","weight":20},{"key":"language_information","label":"语言与信息处理","weight":20},{"key":"multimodal_understanding","label":"多模态理解","weight":10}],"interpretation":"A site-defined relative index of available published evidence; coverage differs between models; not an evaluation run by this website","minimum_benchmark_families":3,"minimum_capability_groups":2,"normalization":"100*(worse_peers+0.5*other_tied_peers)/(peer_count-1)","reference_population":"One frozen recent-model population per current benchmark version and comparable protocol; distinct model versions, no historical peers","rounding":"Exact rational arithmetic; half-up two decimals; equal published scores share rank","source_precedence":["benchmark_owner_or_reviewed_evaluator","own_official_release","other_mainstream_official_comparison"],"version":"auratracer-public-composite-v9","versions":"Latest effective versions established by benchmark maintainer or evaluator","weighting":"Equal benchmark families within each covered capability group; covered base weights proportionally normalized","window_months":6},"observation_count":1272,"pool_revision":"932667e453e157491b3ac058c654ccf8b07806197f90020e06a07f57fba5f4b1","ranked":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":153,"numerator":15100},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench-Science","row_ids":["416c079f92e7711d76f35d3a6da3f408"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"HLE","row_ids":["2541b171f3a9b70175c5eff4ce5a2a47"],"score":100.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4900},"family":"CritPt","row_ids":["29eed13fef0ca4e5917825122c33b6b0"],"score":96.08}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","peer_count":64,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","peer_count":52,"percentile":96.08,"percentile_exact":{"denominator":51,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69}],"score":98.69,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":3564,"numerator":321485},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7400},"family":"GDPval-AA","row_ids":["5c7968ea3f6404bb3d978285ac42aa05"],"score":91.36},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":375},"family":"Terminal-Bench","row_ids":["2db4ed4fb97d84958b935e83df5db43d"],"score":93.75},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["1b1f1f3e062850814bcd282684d4e879"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":1000},"family":"AA-Briefcase","row_ids":["fe408037e14234a002d5a2d7a3b186f0"],"score":90.91},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":75},"family":"OSWorld","row_ids":["f7b8bce78b6bbb414e44451ecece6456"],"score":75.0}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","peer_count":82,"percentile":91.36,"percentile_exact":{"denominator":81,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","peer_count":49,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":78,"percentile":90.91,"percentile_exact":{"denominator":11,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5}],"score":90.2,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":4400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4400},"family":"FrontierCode","row_ids":["54642e364cbdcff99c9cf0483b73a212"],"score":89.8}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","peer_count":50,"percentile":89.8,"percentile_exact":{"denominator":49,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061}],"score":89.8,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":94350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1750},"family":"GDP.pdf","row_ids":["2ef7660125e00e7ecf9ebbabeb38c644"],"score":92.11},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":5050},"family":"AA-Omniscience","row_ids":["00d59d147ff0b7cd3c12bdcaa9a4b516","91da142d2c0422108afa80f33bfb9bb4"],"score":95.28}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339}],"score":93.69,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":100},"family":"MMMU-Pro","row_ids":["34112c44d835a7ef36f154c3c42aa2b1"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["98879b4fa09dc7b005731381bea066e0"],"score":100.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5}],"score":100.0,"weight":10}],"eligible":true,"id":"m-dea86e466cdc70f2","in_window":true,"name":"Gemini 4 Carbon","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":1,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","peer_count":64,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","peer_count":52,"percentile":96.08,"percentile_exact":{"denominator":51,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","peer_count":82,"percentile":91.36,"percentile_exact":{"denominator":81,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","peer_count":49,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":78,"percentile":90.91,"percentile_exact":{"denominator":11,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","peer_count":50,"percentile":89.8,"percentile_exact":{"denominator":49,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5}],"release_date":"2026-10-28","release_status":"published","release_verified":true,"result_updated_at":"2026-10-28","score":93.92,"score_exact":{"denominator":11958374736,"numerator":1123155312275},"source_id":"published/aa:gemini-4-carbon"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":61047,"numerator":5798875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1775},"family":"Terminal-Bench-Science","row_ids":["21af8993ff5f3e38ecc3caf66271b208"],"score":93.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":6200},"family":"HLE","row_ids":["3522855b1814574fd5b3733b32b77769"],"score":98.41},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4750},"family":"CritPt","row_ids":["46e4906d019b793c19b1d443cd35b178"],"score":93.14}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","peer_count":64,"percentile":98.41,"percentile_exact":{"denominator":63,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68}],"score":94.99,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":302841,"numerator":28338056},"families":[{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":100},"family":"GDPval-AA","row_ids":["b44123b055ba24f02daafca81efcb87b"],"score":100.0},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":96},"family":"Terminal-Bench","row_ids":["322131c4aa07af285df46f472061072d"],"score":96.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1750},"family":"AutomationBench","row_ids":["f100a726506576804779fdbd76208d6e"],"score":92.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7600},"family":"AA-Briefcase","row_ids":["5c16c97705a8f70907dc1ba381949448"],"score":98.7},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":250},"family":"OSWorld","row_ids":["cb71933d15a49daa0088c6b64854af7f"],"score":83.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2100},"family":"AA-AnalystAgent","row_ids":["1a02ab6144866fd1b960f19d8fd21513"],"score":91.3}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","peer_count":82,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"4.0","peer_count":76,"percentile":96.0,"percentile_exact":{"denominator":1,"numerator":96},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":78,"percentile":98.7,"percentile_exact":{"denominator":77,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.1","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","peer_count":24,"percentile":91.3,"percentile_exact":{"denominator":23,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625}],"score":93.57,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":1275},"families":[{"benchmark_version":"2.0","exact":{"denominator":1,"numerator":100},"family":"FrontierCode","row_ids":["a1969d9eb66e2bed1a98e36808c17360"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1250},"family":"Vals Code Migration","row_ids":["d2e15e2796cf44b3aabbae2da4bd5b3f"],"score":96.15}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.0","peer_count":50,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"1.1","peer_count":27,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01}],"score":98.08,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2822875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3425},"family":"GDP.pdf","row_ids":["77cd16d89e49025a3285c182b2da553d"],"score":90.13},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1550},"family":"SimpleQA Verified","row_ids":["144b27b350bf50d4c19f78baa94c183a"],"score":91.18},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3500},"family":"AA-Omniscience","row_ids":["3247fbdae3f488e2395593b52951cc4b","035a4d676bedc8d21713e0b6aa5bc0ad"],"score":66.04}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227}],"score":82.45,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":459,"numerator":37630},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"Humanity Sixth Sense","row_ids":["78db71948ffd6a9f7a24a59779294d90"],"score":88.89},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1650},"family":"MMMU-Pro","row_ids":["3dabe34a660554a2d248f440d9ae144c"],"score":97.06},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":60},"family":"Chartography","row_ids":["d2e4430a89fa4e3972bc81f62bbdad07"],"score":60.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","peer_count":28,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":60.0,"percentile_exact":{"denominator":1,"numerator":60},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4}],"score":81.98,"weight":10}],"eligible":true,"id":"m-cfe0652e9d685783","in_window":true,"name":"Claude Opus 5.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":2,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","peer_count":64,"percentile":98.41,"percentile_exact":{"denominator":63,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","peer_count":82,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"4.0","peer_count":76,"percentile":96.0,"percentile_exact":{"denominator":1,"numerator":96},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":78,"percentile":98.7,"percentile_exact":{"denominator":77,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.1","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","peer_count":24,"percentile":91.3,"percentile_exact":{"denominator":23,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.0","peer_count":50,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"1.1","peer_count":27,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","peer_count":28,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":60.0,"percentile_exact":{"denominator":1,"numerator":60},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":"2026-09-28","score":91.44,"score_exact":{"denominator":42566119596,"numerator":3892435594141},"source_id":"anthropic/claude-opus-5-5"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":3958475},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":100},"family":"FrontierMath","row_ids":["cacd9412ff6ba035ea9e826693b09343"],"score":100.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1850},"family":"Terminal-Bench-Science","row_ids":["77175d98c0074eb53a919fa599931946"],"score":97.37},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5900},"family":"HLE","row_ids":["9165499121e5e198448d4536365d9c16"],"score":93.65},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":5000},"family":"CritPt","row_ids":["96d62eb6c5d675afc73f8103f6dbc627"],"score":98.04}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","peer_count":39,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","peer_count":64,"percentile":93.65,"percentile_exact":{"denominator":63,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","peer_count":52,"percentile":98.04,"percentile_exact":{"denominator":51,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695}],"score":97.26,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4542615,"numerator":363697727},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2000},"family":"GDPval-AA","row_ids":["eaa5c4189ba4b51426be70de02910e04"],"score":74.07},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":272},"family":"Terminal-Bench","row_ids":["d04f5531bb77f209689af5059ec9471d"],"score":90.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1575},"family":"AutomationBench","row_ids":["4bd5c8368df638bd0c064dd06c306d48"],"score":82.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6400},"family":"AA-Briefcase","row_ids":["c7f025daa6b0e96ab52dc7df7e3bb96a"],"score":83.12},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1600},"family":"AA-AnalystAgent","row_ids":["9b9634fd2b2367e076a491de1e4f02e6"],"score":69.57}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","peer_count":82,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","peer_count":76,"percentile":90.67,"percentile_exact":{"denominator":3,"numerator":272},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":78,"percentile":83.12,"percentile_exact":{"denominator":77,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","peer_count":24,"percentile":69.57,"percentile_exact":{"denominator":23,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5}],"score":80.06,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":57425},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4500},"family":"FrontierCode","row_ids":["319a73d6f8fba5d605260942d6b8e601"],"score":91.84},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1150},"family":"Vals Code Migration","row_ids":["e330fa37e52713545b5fec3f612240b8"],"score":88.46}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","peer_count":50,"percentile":91.84,"percentile_exact":{"denominator":49,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","peer_count":27,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32}],"score":90.15,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":12153775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3675},"family":"GDP.pdf","row_ids":["f0c0c3073a678bbcb3f5f855da9adff8"],"score":96.71},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1600},"family":"SimpleQA Verified","row_ids":["509c52385e7c5559b4414a3e4bdb93cc"],"score":94.12},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"MultiChallenge","row_ids":["2a39ee205e56064a7ebeddb7db41c3ee"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3400},"family":"AA-Omniscience","row_ids":["ad20786a83f6b0b25b2f134cb2f6c095","8c01a1619e7c484574ab42025543f92a"],"score":64.15}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075}],"score":88.74,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":42175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2500},"family":"Humanity Sixth Sense","row_ids":["3a0983451c53206a0fa5061a863b8a25"],"score":92.59},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1550},"family":"MMMU-Pro","row_ids":["9a17a04ad8c7b9671f59a85153438a32"],"score":91.18}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","peer_count":28,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335}],"score":91.88,"weight":10}],"eligible":true,"id":"m-0f7beea535abe0d6","in_window":true,"name":"GPT-6.1 Sol","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-29","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":3,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","peer_count":39,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","peer_count":64,"percentile":93.65,"percentile_exact":{"denominator":63,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","peer_count":52,"percentile":98.04,"percentile_exact":{"denominator":51,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","peer_count":82,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","peer_count":76,"percentile":90.67,"percentile_exact":{"denominator":3,"numerator":272},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":78,"percentile":83.12,"percentile_exact":{"denominator":77,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","peer_count":24,"percentile":69.57,"percentile_exact":{"denominator":23,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","peer_count":50,"percentile":91.84,"percentile_exact":{"denominator":49,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","peer_count":27,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","peer_count":28,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335}],"release_date":"2026-09-29","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":89.3,"score_exact":{"denominator":496604728620,"numerator":44346438312619},"source_id":"openai/gpt-6.1-sol"},{"benchmark_count":21,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":100},"family":"FrontierMath","row_ids":["0f175b01de79f7689d7f750dac02d89b"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"GPQA Diamond","row_ids":["c863d641570942861efdba9c19093efa"],"score":100.0},{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench-Science","row_ids":["33b37c700e39b0d0c6a88f59bf14c960"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"HLE","row_ids":["955cc9868301478204da3b4815a3ad3c"],"score":100.0},{"benchmark_version":"2 Hard","exact":{"denominator":1,"numerator":100},"family":"CritPt","row_ids":["0a974b1036bafb40411da54c5245579a"],"score":100.0}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","peer_count":39,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","peer_count":52,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705}],"score":100.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":248026779,"numerator":19458728500},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1100},"family":"tau3-Banking","row_ids":["388b55b5431573f02c7a1415569151aa"],"score":84.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5800},"family":"GDPval-AA","row_ids":["b076d23f5b3bf4c32b53727e95233455"],"score":71.6},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":275},"family":"Terminal-Bench","row_ids":["28b064c611914ab9c6c8b9c66db69c4f"],"score":91.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1725},"family":"AutomationBench","row_ids":["fb60ba99063f49f876c4c8d56120f0f3"],"score":90.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6500},"family":"AA-Briefcase","row_ids":["e3f8cbf64c84e82bb4a86c035247a89a"],"score":84.42},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":50},"family":"OSWorld","row_ids":["2ed0ffc858d7ba150a21822c243ebff1"],"score":50.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1750},"family":"AA-AnalystAgent","row_ids":["15975e3d42ab5d82a67bcbe277dd1b83"],"score":76.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","peer_count":82,"percentile":71.6,"percentile_exact":{"denominator":81,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","peer_count":49,"percentile":91.67,"percentile_exact":{"denominator":3,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":90.79,"percentile_exact":{"denominator":19,"numerator":1725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":78,"percentile":84.42,"percentile_exact":{"denominator":77,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125}],"score":78.45,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":1911,"numerator":184900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4800},"family":"FrontierCode","row_ids":["18ca4c50e781eec01bc5749fbb189315"],"score":97.96},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1200},"family":"Vals Code Migration","row_ids":["bc107d6862df5282c7cf8b968206739b"],"score":92.31},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["948d62640ac1b6f74166a175a86cf6a7"],"score":100.0}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","peer_count":50,"percentile":97.96,"percentile_exact":{"denominator":49,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":27,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":9,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1}],"score":96.76,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2954325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3675},"family":"GDP.pdf","row_ids":["96f5c4d98f214c070d53868688e0a1a3"],"score":96.71},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1650},"family":"SimpleQA Verified","row_ids":["3557cefe053ef3a2ca1c95595065295f"],"score":97.06},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3450},"family":"AA-Omniscience","row_ids":["eec9e765997481a4684eb7ff4710087d","6e1af818db0c40a97c73d45f3dd92647"],"score":65.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958}],"score":86.29,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":1377,"numerator":110350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2600},"family":"Humanity Sixth Sense","row_ids":["35c55ac70a4db5208200b7b35fe696c9"],"score":96.3},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1600},"family":"MMMU-Pro","row_ids":["ec04ac79e57d7b9ceb9bc5c32ee3fd30"],"score":94.12},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"Chartography","row_ids":["600568174d3d7470e372c6b4483b5633"],"score":50.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","peer_count":28,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0}],"score":80.14,"weight":10}],"eligible":true,"id":"m-5553b7ca5f3e748a","in_window":true,"name":"GPT-6 Astra","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":4,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","peer_count":39,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","peer_count":52,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","peer_count":82,"percentile":71.6,"percentile_exact":{"denominator":81,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","peer_count":49,"percentile":91.67,"percentile_exact":{"denominator":3,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":90.79,"percentile_exact":{"denominator":19,"numerator":1725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":78,"percentile":84.42,"percentile_exact":{"denominator":77,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","peer_count":50,"percentile":97.96,"percentile_exact":{"denominator":49,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":27,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":9,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","peer_count":28,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":"2026-10-28","score":89.24,"score_exact":{"denominator":21283059798,"numerator":1899215940095},"source_id":"openai/gpt-6-astra"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":5967,"numerator":520600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["0caa25aeefa10afc49082f57afb1971b"],"score":84.62},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":800},"family":"HLE","row_ids":["4e26762c5adc39cb8336491e6db427a8"],"score":88.89},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1500},"family":"CritPt","row_ids":["f2c7d3b6f687d90b18b5ecae64664442"],"score":88.24}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","peer_count":64,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","peer_count":52,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565}],"score":87.25,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3792096,"numerator":357306625},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7900},"family":"GDPval-AA","row_ids":["fa97f8f99150ab638ce3d789e781973c"],"score":97.53},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":2125},"family":"Terminal-Bench","row_ids":["07f1cdf7bbcb0da17ca1fb2ddf289fbe"],"score":88.54},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1775},"family":"AutomationBench","row_ids":["9d53e1ef4d2e9c319b7f0dacff98d83c"],"score":93.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7500},"family":"AA-Briefcase","row_ids":["4f4f6c3a22e75b41412d7850bc4b9cf5"],"score":97.4}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","peer_count":82,"percentile":97.53,"percentile_exact":{"denominator":81,"numerator":7900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","peer_count":49,"percentile":88.54,"percentile_exact":{"denominator":24,"numerator":2125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":78,"percentile":97.4,"percentile_exact":{"denominator":77,"numerator":7500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32}],"score":94.22,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":637,"numerator":51225},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4300},"family":"FrontierCode","row_ids":["36981a7ad16ac52e29b5d161b5c3a617"],"score":87.76},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":950},"family":"Vals Code Migration","row_ids":["3d507b92bee70e08544b9eb65b20267a"],"score":73.08}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","peer_count":50,"percentile":87.76,"percentile_exact":{"denominator":49,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","peer_count":27,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79}],"score":80.42,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":53,"numerator":4825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"GDP.pdf","row_ids":["b8e5730fd93ac21867d62c37b5d60e40"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4350},"family":"AA-Omniscience","row_ids":["a92d6d21339d4ee97a6fa90820847da1","211e11b4d985ead9c997be1fc30d7961"],"score":82.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418}],"score":91.04,"weight":20}],"eligible":true,"id":"m-21b38c9ea818aebe","in_window":true,"name":"Muse Wonder Full","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":5,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","peer_count":64,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","peer_count":52,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","peer_count":82,"percentile":97.53,"percentile_exact":{"denominator":81,"numerator":7900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","peer_count":49,"percentile":88.54,"percentile_exact":{"denominator":24,"numerator":2125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":78,"percentile":97.4,"percentile_exact":{"denominator":77,"numerator":7500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","peer_count":50,"percentile":87.76,"percentile_exact":{"denominator":49,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","peer_count":27,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418}],"release_date":"2026-10-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":88.51,"score_exact":{"denominator":5596519376448,"numerator":495344146904675},"source_id":"meta/muse-wonder-full"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":46512,"numerator":3838925},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":700},"family":"FrontierMath","row_ids":["a0152968ddf4e79954f370988a3aa877"],"score":77.78},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1775},"family":"Terminal-Bench-Science","row_ids":["0fd84fb90c507070cc93b2671c1dacd9"],"score":93.42},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":275},"family":"HLE","row_ids":["36082ffe886fbeb2647876856cf1e5c0"],"score":68.75},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4600},"family":"CritPt","row_ids":["57f29ac78c3749b38fa2eef65ef43c04"],"score":90.2}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","peer_count":52,"percentile":90.2,"percentile_exact":{"denominator":51,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59}],"score":82.54,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":35397,"numerator":3458243},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2600},"family":"GDPval-AA","row_ids":["c0ede537ba74a003e6de650e7903fd9e"],"score":96.3},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":292},"family":"Terminal-Bench","row_ids":["620eabc34583900c0a5610d2452e8c0e"],"score":97.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1800},"family":"AutomationBench","row_ids":["7e4dabafdeb0615d7fe8c2c4159042bd"],"score":94.74},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"AA-Briefcase","row_ids":["609c8649b3a04b44b0d6807fdfe15427"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":100},"family":"OSWorld","row_ids":["e0c59ec207d44193213f25c93cf18f5b"],"score":100.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2250},"family":"AA-AnalystAgent","row_ids":["b14e1b2737d16e7bb80a947abb8a320a"],"score":97.83}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","peer_count":82,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"4.0","peer_count":76,"percentile":97.33,"percentile_exact":{"denominator":3,"numerator":292},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":78,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"score":97.7,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":49,"numerator":4750},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4600},"family":"FrontierCode","row_ids":["2bdac0e90d63fa507e78af95b1c832a5"],"score":93.88},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"Vals Code Migration","row_ids":["165b56e6a0826f1ed0dc82b37b0885fe"],"score":100.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.0","peer_count":50,"percentile":93.88,"percentile_exact":{"denominator":49,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"1.1","peer_count":27,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88}],"score":96.94,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":3278075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1675},"family":"GDP.pdf","row_ids":["7992d55958a238d712427a450a500af9"],"score":88.16},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":650},"family":"SimpleQA Verified","row_ids":["b5dfd8ffb656e8526bed64b86b168f2a"],"score":38.24},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3450},"family":"AA-Omniscience","row_ids":["911853403e469f557e0c742924a5bacb","553708168125ea365f4f1a89df474d93"],"score":65.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923}],"score":63.83,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["34ec823cb4bb85ff7965822a48a3f4ce"],"score":100.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6}],"score":100.0,"weight":10}],"eligible":true,"id":"m-d123610543e5f5ed","in_window":true,"name":"Claude Sonnet 5.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":6,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","peer_count":52,"percentile":90.2,"percentile_exact":{"denominator":51,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","peer_count":82,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"4.0","peer_count":76,"percentile":97.33,"percentile_exact":{"denominator":3,"numerator":292},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":78,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.0","peer_count":50,"percentile":93.88,"percentile_exact":{"denominator":49,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"1.1","peer_count":27,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6}],"release_date":"2026-09-28","release_status":"published","release_verified":true,"result_updated_at":"2026-10-07","score":87.21,"score_exact":{"denominator":100015497792,"numerator":8722585306847},"source_id":"anthropic/claude-sonnet-5-5"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":15912,"numerator":1118525},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":500},"family":"FrontierMath","row_ids":["a21e66eb62f2bfca72e7c625965b008a"],"score":55.56},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1050},"family":"Terminal-Bench-Science","row_ids":["161dc1424fa043da938a084c4e822aa1"],"score":80.77},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":125},"family":"HLE","row_ids":["c73228fc930f557212149ca3c921d3fb"],"score":62.5},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1400},"family":"CritPt","row_ids":["024c0dcfccf03a6bb518b6c682477674"],"score":82.35}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"0.1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Diamond","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","peer_count":52,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505}],"score":70.29,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":21649031},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7700},"family":"GDPval-AA","row_ids":["5449b5fbd439528611d565a71ad28a45"],"score":95.06},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":262},"family":"Terminal-Bench","row_ids":["6fe9193a0e51ae328fa8f19ad476b5c1"],"score":87.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1700},"family":"AutomationBench","row_ids":["b2ddde3d390d97ad479920a7865821c3"],"score":89.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7200},"family":"AA-Briefcase","row_ids":["849c220b415c509e507a4751daf48b0a"],"score":93.51}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2.1","peer_count":82,"percentile":95.06,"percentile_exact":{"denominator":81,"numerator":7700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":78,"percentile":93.51,"percentile_exact":{"denominator":77,"numerator":7200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65}],"score":91.34,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":51000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3700},"family":"FrontierCode","row_ids":["3733f77f03c95d14f61b1419adf05cd6"],"score":75.51},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1100},"family":"Vals Code Migration","row_ids":["ab784b677fcf21bc1b248358effdf278"],"score":84.62}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"2.0","peer_count":50,"percentile":75.51,"percentile_exact":{"denominator":49,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"1.1","peer_count":27,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36}],"score":80.06,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":85700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1800},"family":"GDP.pdf","row_ids":["a4934e52e2688102b1f21d91ca098935"],"score":94.74},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4000},"family":"AA-Omniscience","row_ids":["3f07a1a0073c74e060ddc96a3f9e8bcb","58f09829d0f39ffd8664ffd449b9bfbd"],"score":75.47}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837}],"score":85.1,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":37775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2300},"family":"Humanity Sixth Sense","row_ids":["b9ee5aa138e25de06492fbd597ad313a"],"score":85.19},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1350},"family":"MMMU-Pro","row_ids":["bc79e5409c84a79a9897609e9d7ac22e"],"score":79.41}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","peer_count":28,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","peer_count":35,"percentile":79.41,"percentile_exact":{"denominator":17,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341}],"score":82.3,"weight":10}],"eligible":true,"id":"m-68480fc0423c5034","in_window":true,"name":"Muse Wonder","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":7,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"0.1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Diamond","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","peer_count":52,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2.1","peer_count":82,"percentile":95.06,"percentile_exact":{"denominator":81,"numerator":7700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":78,"percentile":93.51,"percentile_exact":{"denominator":77,"numerator":7200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"2.0","peer_count":50,"percentile":75.51,"percentile_exact":{"denominator":49,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"1.1","peer_count":27,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","peer_count":28,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","peer_count":35,"percentile":79.41,"percentile_exact":{"denominator":17,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341}],"release_date":"2026-10-12","release_status":"published","release_verified":true,"result_updated_at":"2026-10-12","score":81.67,"score_exact":{"denominator":310917743136,"numerator":25393529124989},"source_id":"meta/muse-wonder-1.0"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":41769,"numerator":3441925},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1075},"family":"Terminal-Bench-Science","row_ids":["440fdb6baf103b127060630dee151d5e"],"score":82.69},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5300},"family":"HLE","row_ids":["8cb18c97abf2e38058caf77fed29461c"],"score":84.13},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4100},"family":"CritPt","row_ids":["e4f3938369b9a55c0c1841556cc18184"],"score":80.39}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"0.1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","peer_count":64,"percentile":84.13,"percentile_exact":{"denominator":63,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","peer_count":52,"percentile":80.39,"percentile_exact":{"denominator":51,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495}],"score":82.4,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2725569,"numerator":238699840},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":8000},"family":"GDPval-AA","row_ids":["948bda542167bd107b882e3640f83a0c"],"score":98.77},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench","row_ids":["1c21aa90c01783fcd785c617ba4d90f9"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1850},"family":"AutomationBench","row_ids":["97fcee40bdfab50663d8d9aa18bd9131"],"score":97.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7400},"family":"AA-Briefcase","row_ids":["2b512abdb78eff5062c8cbfe38f0a197"],"score":96.1},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1050},"family":"AA-AnalystAgent","row_ids":["8b93841abeaaa5b29864bb557e6a9da7"],"score":45.65}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","peer_count":82,"percentile":98.77,"percentile_exact":{"denominator":81,"numerator":8000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"4.0","peer_count":76,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":78,"percentile":96.1,"percentile_exact":{"denominator":77,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"score":87.58,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":7,"numerator":500},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":500},"family":"FrontierCode","row_ids":["2cd7a142e0f06d143bb474c89fce8084"],"score":71.43}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"2.0","peer_count":50,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444}],"score":71.43,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":165875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1875},"family":"GDP.pdf","row_ids":["24d56f9893e1402bb483f6bef166ba08"],"score":98.68},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3500},"family":"AA-Omniscience","row_ids":["e4c0067c2ce56fc8c2df02c769685194","faf92c6808beb316ccdbea4a34863bce"],"score":66.04}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008}],"score":82.36,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1400},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1400},"family":"MMMU-Pro","row_ids":["75435cf97a5f0d98b110b2544004171f"],"score":82.35}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422}],"score":82.35,"weight":10}],"eligible":true,"id":"m-66c1aae520b7679a","in_window":true,"name":"Kimi K3.1 Pro","official_evidence":{"artifact_sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","basis":"official_release_page","date":"2026-10-24","sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","url":"https://www.kimi.com/en/blog/"},"primary_profile_id":"native","profile_id":"native","rank":8,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"0.1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","peer_count":64,"percentile":84.13,"percentile_exact":{"denominator":63,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","peer_count":52,"percentile":80.39,"percentile_exact":{"denominator":51,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","peer_count":82,"percentile":98.77,"percentile_exact":{"denominator":81,"numerator":8000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"4.0","peer_count":76,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":78,"percentile":96.1,"percentile_exact":{"denominator":77,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"2.0","peer_count":50,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422}],"release_date":"2026-10-24","release_status":"published","release_verified":true,"result_updated_at":null,"score":81.49,"score_exact":{"denominator":127698358788,"numerator":10405967280655},"source_id":"moonshotai/kimi-k3.1-pro"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":3393175},"families":[{"benchmark_version":"2","exact":{"denominator":3,"numerator":200},"family":"FrontierMath","row_ids":["1004f40613fd7110bf20d6c47c5dcedb"],"score":66.67},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1700},"family":"Terminal-Bench-Science","row_ids":["50ddd51f0edebe68ad08c74703710c38"],"score":89.47},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5800},"family":"HLE","row_ids":["b185a4fb3502b3f94e9e93582e0d93a3"],"score":92.06},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1450},"family":"CritPt","row_ids":["3130f94996ac2afcc2f6ea8e78d8c87e"],"score":85.29}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","peer_count":39,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","peer_count":64,"percentile":92.06,"percentile_exact":{"denominator":63,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":83.37,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":16353414,"numerator":1123315619},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6500},"family":"GDPval-AA","row_ids":["feff6b1ffb0f6cb8ff2773defec53d29"],"score":80.25},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":278},"family":"Terminal-Bench","row_ids":["58499c853c2644c7eb9a4bc16fc208c6"],"score":92.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1875},"family":"AutomationBench","row_ids":["a2113bd24aca42a48eba4972dbb6e6d4"],"score":98.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5800},"family":"AA-Briefcase","row_ids":["782453050df6541e218363a18f9bb044"],"score":75.32},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"OSWorld","row_ids":["f61750e8ba74bd2c7928dd7c56124cb4"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1500},"family":"AA-AnalystAgent","row_ids":["4f1262f25c1f33cd755aebe2e08018dd"],"score":65.22}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","peer_count":82,"percentile":80.25,"percentile_exact":{"denominator":81,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","peer_count":76,"percentile":92.67,"percentile_exact":{"denominator":3,"numerator":278},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":78,"percentile":75.32,"percentile_exact":{"denominator":77,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","peer_count":24,"percentile":65.22,"percentile_exact":{"denominator":23,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875}],"score":68.69,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":1911,"numerator":161950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3600},"family":"FrontierCode","row_ids":["aabef6305c7bf9e588cc8a4221b8394d"],"score":73.47},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1050},"family":"Vals Code Migration","row_ids":["2905dc672ed8db458e1dc8c5a09c04a7"],"score":80.77},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":100},"family":"MirrorCode","row_ids":["d2afdbf96101f2fb4499bd4ee1fc2881"],"score":100.0}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","peer_count":50,"percentile":73.47,"percentile_exact":{"denominator":49,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","peer_count":27,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889}],"score":84.75,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":2014,"numerator":175925},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1575},"family":"GDP.pdf","row_ids":["da4cc7b10c537dba69aca5448e60f8f7"],"score":82.89},{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":100},"family":"SimpleQA Verified","row_ids":["56d131d2a5cbdffe166bdda67e598654"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"MultiChallenge","row_ids":["2db7888810c9c42969f9152c713f9d6e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4850},"family":"AA-Omniscience","row_ids":["0588c4d7908c2d87deb09a5a03cbe1c3","2f9d0f30b6d18925324be8097217f20f"],"score":91.51}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383}],"score":87.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":2,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["18175febe04f7246f991614fc6ca93f9"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"Chartography","row_ids":["13ad1c51a1b24d02661158660b44e5e0"],"score":75.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","peer_count":28,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6}],"score":87.5,"weight":10}],"eligible":true,"id":"m-7ff6a9f995dcaa1c","in_window":true,"name":"Gemini 4 Argon","official_evidence":{"artifact_sha256":"555ea742fc3bcfcd343ab416bbcdf209448d15c9f06f0240eb19c9875fdccf00","basis":"official_release_page","date":"2026-10-13","sha256":"555ea742fc3bcfcd343ab416bbcdf209448d15c9f06f0240eb19c9875fdccf00","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-argon/"},"primary_profile_id":"native","profile_id":"native","rank":9,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","peer_count":39,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","peer_count":64,"percentile":92.06,"percentile_exact":{"denominator":63,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","peer_count":82,"percentile":80.25,"percentile_exact":{"denominator":81,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","peer_count":76,"percentile":92.67,"percentile_exact":{"denominator":3,"numerator":278},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":78,"percentile":75.32,"percentile_exact":{"denominator":77,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","peer_count":24,"percentile":65.22,"percentile_exact":{"denominator":23,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","peer_count":50,"percentile":73.47,"percentile_exact":{"denominator":49,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","peer_count":27,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","peer_count":28,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6}],"release_date":"2026-10-13","release_status":"published","release_verified":true,"result_updated_at":"2026-10-28","score":81.19,"score_exact":{"denominator":2681665534548,"numerator":217712535238087},"source_id":"google/gemini-4-argon"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":31824,"numerator":2630125},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":800},"family":"FrontierMath","row_ids":["428bcb6f2312f54ea131660ec5510f0c"],"score":88.89},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1200},"family":"Terminal-Bench-Science","row_ids":["3a64ac33a8e0bec7c9bbb47064e6200b"],"score":92.31},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":225},"family":"HLE","row_ids":["9c15f08ff3f1d95aac1ab4460a6a090b"],"score":56.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4750},"family":"CritPt","row_ids":["624fa427f469bd9c8115ca8491b3f8a8"],"score":93.14}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","peer_count":53,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68}],"score":82.65,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":17835809},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5600},"family":"GDPval-AA","row_ids":["b0fdd931f8b81700b478f4129dec593b"],"score":69.14},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":268},"family":"Terminal-Bench","row_ids":["35604c8a1315056deeba470a2491f933"],"score":89.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1450},"family":"AutomationBench","row_ids":["1b6722ff89e791988385d9c44034ffe3"],"score":76.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5100},"family":"AA-Briefcase","row_ids":["07c0ef81ba075e768f25e1177f4ce283"],"score":66.23}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","peer_count":82,"percentile":69.14,"percentile_exact":{"denominator":81,"numerator":5600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","peer_count":76,"percentile":89.33,"percentile_exact":{"denominator":3,"numerator":268},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":78,"percentile":66.23,"percentile_exact":{"denominator":77,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36}],"score":75.25,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":4700},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4700},"family":"FrontierCode","row_ids":["bf241286cb096d94e2aa92889567c092"],"score":95.92}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","peer_count":50,"percentile":95.92,"percentile_exact":{"denominator":49,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505}],"score":95.92,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":9124175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1775},"family":"GDP.pdf","row_ids":["5a11c32835ae26c4dc93dd63135b1a2a"],"score":93.42},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1150},"family":"SimpleQA Verified","row_ids":["52a58ef3bf54eb42e218e3c094aec598"],"score":67.65},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"MultiChallenge","row_ids":["c65e268d91030372566c3ec5c7a1ab64"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3600},"family":"AA-Omniscience","row_ids":["43cafc16779c4f9a017fe8ae9ccdcb81","5ac877348d0fb5ad33d4ed61b4154929"],"score":67.92}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129}],"score":66.62,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1450},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1450},"family":"MMMU-Pro","row_ids":["cd4afdb9196fc47218c2c1085364d08a"],"score":85.29}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555}],"score":85.29,"weight":10}],"eligible":true,"id":"m-1dcd13c2fce7d97b","in_window":true,"name":"GPT-6.1 Terra","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":10,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","peer_count":53,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","peer_count":82,"percentile":69.14,"percentile_exact":{"denominator":81,"numerator":5600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","peer_count":76,"percentile":89.33,"percentile_exact":{"denominator":3,"numerator":268},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":78,"percentile":66.23,"percentile_exact":{"denominator":77,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","peer_count":50,"percentile":95.92,"percentile_exact":{"denominator":49,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555}],"release_date":"2026-10-26","release_status":"published","release_verified":true,"result_updated_at":"2026-10-26","score":80.51,"score_exact":{"denominator":621835486272,"numerator":50065762899577},"source_id":"openai/gpt-6.1-terra"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":533596,"numerator":44974135},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":500},"family":"FrontierMath","row_ids":["a450c91f279ff648cb2b208a74552ba0"],"score":71.43},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5700},"family":"GPQA Diamond","row_ids":["0539c349124105a728b2ffaeadc2dabd"],"score":96.61},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1650},"family":"Terminal-Bench-Science","row_ids":["2012b47ec868b52e455d3d8d66a1a48a"],"score":86.84},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":325},"family":"HLE","row_ids":["16b64e0ecfeb64ebd3bc40ba58f5f691"],"score":81.25},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1450},"family":"CritPt","row_ids":["2a25d966a3bebfa161181074ba7fbad0"],"score":85.29}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":96.61,"percentile_exact":{"denominator":59,"numerator":5700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","peer_count":39,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":84.28,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":496053558,"numerator":42907839551},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":2375},"family":"tau3-Banking","row_ids":["a5ec26f61468360fa82a5667cce554db"],"score":91.35},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7600},"family":"GDPval-AA","row_ids":["1bc96450387b9110981d343f4dfd20eb"],"score":93.83},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":262},"family":"Terminal-Bench","row_ids":["71f828c6aa1f518ef08599ca0a1b0570"],"score":87.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1400},"family":"AutomationBench","row_ids":["9a8fd23ce0ae510627402be3083ec808"],"score":73.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7300},"family":"AA-Briefcase","row_ids":["cab25ff85e887187f7f900705800483c"],"score":94.81},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"OSWorld","row_ids":["520b3d1c3a49b9ec75dd20571c69ae96"],"score":66.67},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2250},"family":"AA-AnalystAgent","row_ids":["172f756dc0c16d9248bcf51116651a4b"],"score":97.83}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","peer_count":82,"percentile":93.83,"percentile_exact":{"denominator":81,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":78,"percentile":94.81,"percentile_exact":{"denominator":77,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.1","peer_count":7,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"score":86.5,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":98,"numerator":5325},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4100},"family":"FrontierCode","row_ids":["468e0160c503684573c12f005950e313"],"score":83.67},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["a9df38e2550ef9f8d566441fba3e6c66"],"score":25.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.0","peer_count":50,"percentile":83.67,"percentile_exact":{"denominator":49,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"1.1","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4}],"score":54.34,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":102714,"numerator":8141825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3425},"family":"GDP.pdf","row_ids":["5d9e8383609a67f5871ab595963f24e2"],"score":90.13},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1500},"family":"SimpleQA Verified","row_ids":["add598a99258de2861924fc25c825623"],"score":88.24},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3150},"family":"AA-Omniscience","row_ids":["b829940da89808936844666d294a78aa","2cbd1e01e64dad121dca7aaf300f12be"],"score":59.43}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996}],"score":79.27,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":9,"numerator":800},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":700},"family":"Humanity Sixth Sense","row_ids":["2d80b1607b40fc4d214cb22e259d1221"],"score":77.78},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["a7fcafd577ada1f13084bdc5d25fc8ab"],"score":100.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","peer_count":28,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4}],"score":88.89,"weight":10}],"eligible":true,"id":"m-98ee7005cf45b532","in_window":true,"name":"Claude Fable 5.1","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":11,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":96.61,"percentile_exact":{"denominator":59,"numerator":5700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","peer_count":39,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","peer_count":82,"percentile":93.83,"percentile_exact":{"denominator":81,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":78,"percentile":94.81,"percentile_exact":{"denominator":77,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.1","peer_count":7,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.0","peer_count":50,"percentile":83.67,"percentile_exact":{"denominator":49,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"1.1","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","peer_count":28,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4}],"release_date":"2026-09-01","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":78.31,"score_exact":{"denominator":30136812673968,"numerator":2359877379520649},"source_id":"anthropic/claude-fable-5-1"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3717,"numerator":264100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5200},"family":"GPQA Diamond","row_ids":["c6f5e59b8c457f8fd73cf7fe23e60786"],"score":88.14},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3400},"family":"HLE","row_ids":["1e856649fb09c4da785f8c4ea8ad05e0"],"score":53.97}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":88.14,"percentile_exact":{"denominator":59,"numerator":5200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond","peer_count":64,"percentile":53.97,"percentile_exact":{"denominator":63,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163}],"score":71.05,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":603860552},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1150},"family":"tau3-Banking","row_ids":["24d253e4fe3d6de3cc614f0a892166b1"],"score":88.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6700},"family":"GDPval-AA","row_ids":["c47c79a0ac9d03dc0858a3e32eb10cf4"],"score":82.72},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":68},"family":"Terminal-Bench","row_ids":["0978fbc73a68e2ee70d02c130e0d1cff"],"score":68.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1250},"family":"AutomationBench","row_ids":["12fa4c01c24c67c3a86dd523987aed83"],"score":65.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6700},"family":"AA-Briefcase","row_ids":["62ba5fa448ec9e7825ee5093f17e487d"],"score":87.01}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v2.1","peer_count":82,"percentile":82.72,"percentile_exact":{"denominator":81,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"4.0","peer_count":76,"percentile":68.0,"percentile_exact":{"denominator":1,"numerator":68},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":78,"percentile":87.01,"percentile_exact":{"denominator":77,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51}],"score":78.4,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["80a064d01bdbcc1f636799e9dd23518a"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"1.1","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7}],"score":100.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":6042,"numerator":401725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["1bcc725977832eb3960634353511e9be"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2525},"family":"GDP.pdf","row_ids":["046ca38af4c5ad3c8241882797259027"],"score":66.45},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1750},"family":"AA-Omniscience","row_ids":["0351ad1258b35881ec0b312bb72c8d29","8283a5fc47f30f74b98c14b10fde0469"],"score":33.02}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129}],"score":66.49,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1050},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1050},"family":"MMMU-Pro","row_ids":["42c55648590ad4a69cdc2a21de6faac7"],"score":61.76}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676}],"score":61.76,"weight":10}],"eligible":true,"id":"m-e1cba225d1430824","in_window":true,"name":"Qwen3.8-Flash-Next","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":12,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":88.14,"percentile_exact":{"denominator":59,"numerator":5200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond","peer_count":64,"percentile":53.97,"percentile_exact":{"denominator":63,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v2.1","peer_count":82,"percentile":82.72,"percentile_exact":{"denominator":81,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"4.0","peer_count":76,"percentile":68.0,"percentile_exact":{"denominator":1,"numerator":68},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":78,"percentile":87.01,"percentile_exact":{"denominator":77,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"1.1","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"result_updated_at":"2026-08-26","score":76.84,"score_exact":{"denominator":818935127010,"numerator":62923865665709},"source_id":"alibaba/qwen3.8-flash-next"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":27846,"numerator":2129875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1975},"family":"Terminal-Bench-Science","row_ids":["bb1f3609ad48f7662af1e4f3a15cef7a"],"score":75.96},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1700},"family":"HLE","row_ids":["5198f9fefccca8b5961dab3ccd43cfb2"],"score":80.95},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3700},"family":"CritPt","row_ids":["6c2b0cbfd16faf311bd4edca68e40d0e"],"score":72.55}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","peer_count":64,"percentile":80.95,"percentile_exact":{"denominator":21,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","peer_count":52,"percentile":72.55,"percentile_exact":{"denominator":51,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39}],"score":76.49,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1197,"numerator":95609},"families":[{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":248},"family":"Terminal-Bench","row_ids":["b9ebef353db3bd587384633039b7a5bb"],"score":82.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1625},"family":"AutomationBench","row_ids":["01aaf3c58f06d7a8962fe3fd60fa4286"],"score":85.53},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":500},"family":"AA-Briefcase","row_ids":["732f112b308243fb337093d80efff0c3"],"score":71.43}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","peer_count":76,"percentile":82.67,"percentile_exact":{"denominator":3,"numerator":248},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":85.53,"percentile_exact":{"denominator":19,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":78,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69}],"score":79.87,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":4000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4000},"family":"FrontierCode","row_ids":["8ed6fd9401411cecf6b0e548f78e12a5"],"score":81.63}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","peer_count":50,"percentile":81.63,"percentile_exact":{"denominator":49,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101}],"score":81.63,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1007,"numerator":64350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1550},"family":"GDP.pdf","row_ids":["b84d512b0d7ca681d5568eba0c2a9675"],"score":81.58},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2450},"family":"AA-Omniscience","row_ids":["795265b8e61a4f96a23b56d372ad2050","7d74696c1e12470b21a043973afd962c"],"score":46.23}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356}],"score":63.9,"weight":20}],"eligible":true,"id":"m-b88efd1e16a0b031","in_window":true,"name":"Tencent Hy4","official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-11-03","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":13,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","peer_count":64,"percentile":80.95,"percentile_exact":{"denominator":21,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","peer_count":52,"percentile":72.55,"percentile_exact":{"denominator":51,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","peer_count":76,"percentile":82.67,"percentile_exact":{"denominator":3,"numerator":248},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":85.53,"percentile_exact":{"denominator":19,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":78,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","peer_count":50,"percentile":81.63,"percentile_exact":{"denominator":49,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356}],"release_date":"2026-11-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":75.77,"score_exact":{"denominator":392572908,"numerator":29747186485},"source_id":"published/aa:hunyuan-dev1030-aa"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":294400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5600},"family":"GPQA Diamond","row_ids":["b2e2268e4647b383f382e5a7860dafa3"],"score":94.92},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4000},"family":"HLE","row_ids":["b6295a8854af11f8a1045539e4932330"],"score":63.49}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":94.92,"percentile_exact":{"denominator":59,"numerator":5600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond","peer_count":64,"percentile":63.49,"percentile_exact":{"denominator":63,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201}],"score":79.2,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":10783773,"numerator":844623034},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1275},"family":"tau3-Banking","row_ids":["960dd18162897dbffed3123d8b9728ee"],"score":98.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6200},"family":"GDPval-AA","row_ids":["fcdd405fcadd197dbec16caec46c9124"],"score":76.54},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":56},"family":"Terminal-Bench","row_ids":["8981298b68f7a13ad078d01ac54f2b30"],"score":56.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1325},"family":"AutomationBench","row_ids":["f2b687ab9797dca77a0de360cefeb5cf"],"score":69.74},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5400},"family":"AA-Briefcase","row_ids":["be21483d79b65c8a6ad01a6eee1e0103"],"score":70.13},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"EnterpriseOps-Gym","row_ids":["32758e4ec324e0ec55fbc6f952cf8937"],"score":88.89},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":800},"family":"MCP Atlas","row_ids":["7b2615183b2725bb4693774f30af4a5b"],"score":88.89}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":98.08,"percentile_exact":{"denominator":13,"numerator":1275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v2.1","peer_count":82,"percentile":76.54,"percentile_exact":{"denominator":81,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"4.0","peer_count":76,"percentile":56.0,"percentile_exact":{"denominator":1,"numerator":56},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":78,"percentile":70.13,"percentile_exact":{"denominator":77,"numerator":5400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA run","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5}],"score":78.32,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":238325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2525},"family":"GDP.pdf","row_ids":["93e5cef565694d0b6ece46832b2fc2e8"],"score":66.45},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2750},"family":"AA-Omniscience","row_ids":["872cbb8006efe075a4c9b62b4f118011","7f66b653a79b36bb8c9fbcd69744e5f5"],"score":51.89}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656}],"score":59.17,"weight":20}],"eligible":true,"id":"m-fb920161af1862b7","in_window":true,"name":"Qwen3.8 2.4T A95B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":14,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":94.92,"percentile_exact":{"denominator":59,"numerator":5600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond","peer_count":64,"percentile":63.49,"percentile_exact":{"denominator":63,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":98.08,"percentile_exact":{"denominator":13,"numerator":1275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v2.1","peer_count":82,"percentile":76.54,"percentile_exact":{"denominator":81,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"4.0","peer_count":76,"percentile":56.0,"percentile_exact":{"denominator":1,"numerator":56},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":78,"percentile":70.13,"percentile_exact":{"denominator":77,"numerator":5400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA run","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":73.16,"score_exact":{"denominator":472092014394,"numerator":34540416783815},"source_id":"alibaba/qwen3.8-2.4t-a95b"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":5967,"numerator":413600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":800},"family":"Terminal-Bench-Science","row_ids":["7ead6883bce4c045cf26f5b116b4cd48"],"score":61.54},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":700},"family":"HLE","row_ids":["89d939473fcae282b052740fc3fcfd7a"],"score":77.78},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3500},"family":"CritPt","row_ids":["34dffe7c2a7d437da340dfb5246fe9c8"],"score":68.63}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","peer_count":64,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","peer_count":52,"percentile":68.63,"percentile_exact":{"denominator":51,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"score":69.31,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":35432397,"numerator":2557154263},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1050},"family":"tau3-Banking","row_ids":["66284fe19cdf80bb97a995a814e23753"],"score":80.77},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6400},"family":"GDPval-AA","row_ids":["f53e9f08adf2987f0e2c59d35bf57232"],"score":79.01},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":184},"family":"Terminal-Bench","row_ids":["b334bad67fbd4a96eed8e669e0111fbb"],"score":61.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1025},"family":"AutomationBench","row_ids":["fb9e5b876d81fca5e8cf53e8169709cb"],"score":53.95},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":900},"family":"AA-Briefcase","row_ids":["ae0bf141fcba5d508b0da173dcf01538"],"score":81.82},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":650},"family":"EnterpriseOps-Gym","row_ids":["ccd3047c248ced9697a972d3c394d08d"],"score":72.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1750},"family":"AA-AnalystAgent","row_ids":["e23e94bce569e3668ea49a1736f85e07"],"score":76.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","peer_count":82,"percentile":79.01,"percentile_exact":{"denominator":81,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","peer_count":76,"percentile":61.33,"percentile_exact":{"denominator":3,"numerator":184},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":78,"percentile":81.82,"percentile_exact":{"denominator":11,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125}],"score":72.17,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":3400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3400},"family":"FrontierCode","row_ids":["0fe380f78c53ebf4d268a83df481e0ad"],"score":69.39}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","peer_count":50,"percentile":69.39,"percentile_exact":{"denominator":49,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101}],"score":69.39,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":77450},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1650},"family":"GDP.pdf","row_ids":["792f8618c792ce48ee9c7c6155488bb6"],"score":86.84},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3550},"family":"AA-Omniscience","row_ids":["28a0e72d64e34dbdaef952ff1ff713b4","1a5d81b92d3a8437744af4b44b483171"],"score":66.98}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385}],"score":76.91,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1250},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1250},"family":"MMMU-Pro","row_ids":["e681be1aa8a9a82be96c894925b92cdd"],"score":73.53}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185}],"score":73.53,"weight":10}],"eligible":true,"id":"m-f33f6c88c27b2b9a","in_window":true,"name":"Inkling 2 Small","official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-11-05","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":15,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","peer_count":64,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","peer_count":52,"percentile":68.63,"percentile_exact":{"denominator":51,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","peer_count":82,"percentile":79.01,"percentile_exact":{"denominator":81,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","peer_count":76,"percentile":61.33,"percentile_exact":{"denominator":3,"numerator":184},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":78,"percentile":81.82,"percentile_exact":{"denominator":11,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","peer_count":50,"percentile":69.39,"percentile_exact":{"denominator":49,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185}],"release_date":"2026-11-05","release_status":"published","release_verified":true,"result_updated_at":null,"score":71.98,"score_exact":{"denominator":893888511516,"numerator":64345620454681},"source_id":"published/aa:inkling-small-1102-rc17"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":821457,"numerator":59247200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5350},"family":"GPQA Diamond","row_ids":["65e0dc8726d7a3e7afde224768d43963"],"score":90.68},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":950},"family":"Terminal-Bench-Science","row_ids":["4ea60fe9901ae1fc5d3c1daf559ab6e1"],"score":73.08},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4400},"family":"HLE","row_ids":["2ab50d9a5a2d86ccb853771d429875d3"],"score":69.84},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["b1d4b1cfdd4d38290545201e6c39db63"],"score":54.9}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","included":true,"model_id":"m-825009614c016c1d","original_version_label":"0.1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond","peer_count":64,"percentile":69.84,"percentile_exact":{"denominator":63,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":72.12,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":2267673408,"numerator":186901243525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["ae1b88ab2d0880e90cfd75002e9f4d77"],"score":100.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1225},"family":"tau3-Banking","row_ids":["1cd94c3f93e80676324cc2282ca34aa4"],"score":94.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7100},"family":"GDPval-AA","row_ids":["8fccbdc41b807b2ebbe41cff024b0690"],"score":87.65},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1675},"family":"Terminal-Bench","row_ids":["0d2e487f06bf23cbe7ee7f90f6003d16"],"score":69.79},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1275},"family":"AutomationBench","row_ids":["308f05669cd695ca55d1d348f97d3b83"],"score":67.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6900},"family":"AA-Briefcase","row_ids":["ecc6d7d48519d5a999fc5f71c76c0b6d"],"score":89.61},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":850},"family":"EnterpriseOps-Gym","row_ids":["aaa4039880a2e3c56a0fee65a0cf496e"],"score":94.44},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1300},"family":"AA-AnalystAgent","row_ids":["cee5a16789123af6ef334475ebff73a7"],"score":56.52}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","included":true,"model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":94.23,"percentile_exact":{"denominator":13,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2.1","peer_count":82,"percentile":87.65,"percentile_exact":{"denominator":81,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","included":true,"model_id":"m-825009614c016c1d","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","included":true,"model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":67.11,"percentile_exact":{"denominator":19,"numerator":1275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":78,"percentile":89.61,"percentile_exact":{"denominator":77,"numerator":6900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","peer_count":24,"percentile":56.52,"percentile_exact":{"denominator":23,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45}],"score":82.42,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":40975},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3100},"family":"FrontierCode","row_ids":["6cbe0b4839fdc145889ebffe2def2a50"],"score":63.27},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":850},"family":"Vals Code Migration","row_ids":["62c5b15d1c0217ab796423502244a099"],"score":65.38}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","included":true,"model_id":"m-825009614c016c1d","original_version_label":"2.0","peer_count":50,"percentile":63.27,"percentile_exact":{"denominator":49,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","included":true,"model_id":"m-825009614c016c1d","original_version_label":"1.1","peer_count":27,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19}],"score":64.32,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2127575},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3225},"family":"GDP.pdf","row_ids":["5928f02dfd05733dd4d23f64991980e9"],"score":84.87},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":700},"family":"SimpleQA Verified","row_ids":["dd26bef806a0bba267bbd1b02770b43f"],"score":41.18},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3200},"family":"AA-Omniscience","row_ids":["c42d799db980160b63c3bab31d80fa02","602505b7f91e3456e6da6058f7b120bf"],"score":60.38}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","included":true,"model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088}],"score":62.14,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1300},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1300},"family":"MMMU-Pro","row_ids":["502735168b948696e5ae04c6357e58b6"],"score":76.47}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","peer_count":35,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884}],"score":76.47,"weight":10}],"eligible":true,"id":"m-825009614c016c1d","in_window":true,"name":"Qwen3.8 Max (0902)","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":16,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","included":true,"model_id":"m-825009614c016c1d","original_version_label":"0.1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond","peer_count":64,"percentile":69.84,"percentile_exact":{"denominator":63,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","included":true,"model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":94.23,"percentile_exact":{"denominator":13,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2.1","peer_count":82,"percentile":87.65,"percentile_exact":{"denominator":81,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","included":true,"model_id":"m-825009614c016c1d","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","included":true,"model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":67.11,"percentile_exact":{"denominator":19,"numerator":1275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":78,"percentile":89.61,"percentile_exact":{"denominator":77,"numerator":6900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","peer_count":24,"percentile":56.52,"percentile_exact":{"denominator":23,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","included":true,"model_id":"m-825009614c016c1d","original_version_label":"2.0","peer_count":50,"percentile":63.27,"percentile_exact":{"denominator":49,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","included":true,"model_id":"m-825009614c016c1d","original_version_label":"1.1","peer_count":27,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","included":true,"model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","peer_count":35,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"result_updated_at":"2026-09-02","score":71.58,"score_exact":{"denominator":3375323019484416,"numerator":241593097629493205},"source_id":"alibaba/qwen3.8-max-0902"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":3,"exact":{"denominator":2128,"numerator":133225},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":300},"family":"FrontierMath","row_ids":["88ff5cf09a8ea0336d12c5e7aac222d3"],"score":42.86},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":225},"family":"GPQA Diamond","row_ids":["e80c9d5d961020a2775a86f5a8e8c7ff"],"score":56.25},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1450},"family":"Terminal-Bench-Science","row_ids":["12c1a7b1c4eb18f0409ef269987fa5ca"],"score":76.32},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":75},"family":"HLE","row_ids":["852fa04b612ed916ac880c4c5a9b046b"],"score":75.0}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v2","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","peer_count":39,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6}],"score":62.61,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":6,"exact":{"denominator":252,"numerator":18965},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["803b3c7ebd5d583be94f718268d22790"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"GDPval-AA","row_ids":["3e2aa560fb85d043c269e4880c8d1ed7"],"score":66.67},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":325},"family":"Terminal-Bench","row_ids":["7df1cc76a5ce426695809152ab7c2b62"],"score":81.25},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"AutomationBench","row_ids":["ae96d3d1f5d89c47f0a27de7f5eca5d4"],"score":40.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":800},"family":"AA-Briefcase","row_ids":["a01ae7f3c128573c113df7e946ef1073"],"score":88.89},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":50},"family":"OSWorld","row_ids":["5f71dd3fff3f2fe7b85716b5a982127d"],"score":50.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":100},"family":"MCP Atlas","row_ids":["2ccdf17c2a9e8d9727ba57cd7a67ff5d"],"score":100.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","included":true,"model_id":"m-568e0eb3a95b5113","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","peer_count":49,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":7,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8}],"score":75.26,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":75},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":75},"family":"DeepSWE","row_ids":["d36cdba4dae45a238d9898e2ce34c241"],"score":75.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7}],"score":75.0,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":2,"exact":{"denominator":27,"numerator":1730},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1300},"family":"Humanity Sixth Sense","row_ids":["175d0f5744557d89454896c1420ebf13"],"score":48.15},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":80},"family":"Chartography","row_ids":["1ce2cb7b64555f9b41b3b5999f8e70ba"],"score":80.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","peer_count":28,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4}],"score":64.07,"weight":10}],"eligible":true,"id":"m-568e0eb3a95b5113","in_window":true,"name":"Claude Opus 5","official_evidence":{"artifact_sha256":"2f2269c3e1f9030bf9adc7e06e1979500968e3fc300eedd57a2b54424cab1694","basis":"official_release_page","date":"2026-07-24","sha256":"2f2269c3e1f9030bf9adc7e06e1979500968e3fc300eedd57a2b54424cab1694","url":"https://www.anthropic.com/news/claude-opus-5"},"primary_profile_id":"native","profile_id":"native","rank":17,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v2","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","peer_count":39,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","included":true,"model_id":"m-568e0eb3a95b5113","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","peer_count":49,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":7,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","peer_count":28,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4}],"release_date":"2026-07-24","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":69.84,"score_exact":{"denominator":131328,"numerator":9172165},"source_id":"anthropic/claude-opus-5"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":61047,"numerator":3863750},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["5f766555b014c71ff2d651ccb2fc76d7"],"score":57.89},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4300},"family":"HLE","row_ids":["353c470ec7dd55d7fbbc616285180c31"],"score":68.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3250},"family":"CritPt","row_ids":["539eef88fa72b976d41f8f0e6b20d640"],"score":63.73}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","peer_count":64,"percentile":68.25,"percentile_exact":{"denominator":63,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335}],"score":63.29,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2725569,"numerator":228484706},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7300},"family":"GDPval-AA","row_ids":["4d4ab3fcd8c470a4ca605a1e507ade3e"],"score":90.12},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":70},"family":"Terminal-Bench","row_ids":["a073820ddc97f09d7966e361779cf68e"],"score":70.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1600},"family":"AutomationBench","row_ids":["5dc195401c512a05556701293da42306"],"score":84.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7100},"family":"AA-Briefcase","row_ids":["7ce9f4dd8dbbce093592c7a8603e6868"],"score":92.21},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1900},"family":"AA-AnalystAgent","row_ids":["1c70059543aa3443966891f5d87280d7"],"score":82.61}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","peer_count":82,"percentile":90.12,"percentile_exact":{"denominator":81,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":84.21,"percentile_exact":{"denominator":19,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":78,"percentile":92.21,"percentile_exact":{"denominator":77,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","peer_count":24,"percentile":82.61,"percentile_exact":{"denominator":23,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":83.83,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":43725},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3900},"family":"FrontierCode","row_ids":["efab926973de754a7bc887e6a5732e79"],"score":79.59},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":750},"family":"Vals Code Migration","row_ids":["575f9b1e7b6ccadadc3e13c6241b21d4"],"score":57.69}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"2.0","peer_count":50,"percentile":79.59,"percentile_exact":{"denominator":49,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"1.1","peer_count":27,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22}],"score":68.64,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":901,"numerator":65725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"GDP.pdf","row_ids":["06f207d047d6194d347719cbe681b11e"],"score":75.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1050},"family":"SimpleQA Verified","row_ids":["1bbbdc3f7be54b8c3de3afd5eb04f4c5"],"score":61.76},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4350},"family":"AA-Omniscience","row_ids":["86bdedd227d21984b4d30c6ee2bdc8b4","5789f26e6b16a0b75b183dcbb8cf44bd"],"score":82.08}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947}],"score":72.95,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":700},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":700},"family":"MMMU-Pro","row_ids":["6a15ad1c8de4bbfb578c4f789e53c9b8"],"score":41.18}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341}],"score":41.18,"weight":10}],"eligible":true,"id":"m-7187cf2b01d41935","in_window":true,"name":"Grok 4.7","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":18,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","peer_count":64,"percentile":68.25,"percentile_exact":{"denominator":63,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","peer_count":82,"percentile":90.12,"percentile_exact":{"denominator":81,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":84.21,"percentile_exact":{"denominator":19,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":78,"percentile":92.21,"percentile_exact":{"denominator":77,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","peer_count":24,"percentile":82.61,"percentile_exact":{"denominator":23,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"2.0","peer_count":50,"percentile":79.59,"percentile_exact":{"denominator":49,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"1.1","peer_count":27,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341}],"release_date":"2026-09-21","release_status":"published","release_verified":true,"result_updated_at":"2026-09-22","score":69.22,"score_exact":{"denominator":223472127879,"numerator":15467797601309},"source_id":"xai/grok-4.7"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":114342,"numerator":7985875},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5800},"family":"GPQA Diamond","row_ids":["93a25d2afe298e1250966bfe70f4f70f"],"score":98.31},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1000},"family":"Terminal-Bench-Science","row_ids":["4c2b94e36c85b9d8f2e3d83490a34034"],"score":52.63},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":50},"family":"HLE","row_ids":["08dc502d8d4aeb42de93130ed556d391"],"score":50.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4000},"family":"CritPt","row_ids":["4e088fc1433335f9f212c9f9ffa8b974"],"score":78.43}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":98.31,"percentile_exact":{"denominator":59,"numerator":5800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","peer_count":39,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","peer_count":17,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","peer_count":52,"percentile":78.43,"percentile_exact":{"denominator":51,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465}],"score":69.84,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":5594589,"numerator":380137666},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1125},"family":"tau3-Banking","row_ids":["ceb2e04d7badea9a2869bdbce0586635"],"score":86.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5300},"family":"GDPval-AA","row_ids":["5e4d6ed857d58561b430218e5d2e3d99"],"score":65.43},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":64},"family":"Terminal-Bench","row_ids":["477576b238b66114c394a26ca4a20345"],"score":64.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"AutomationBench","row_ids":["44e1f8cce1da50e102dced9197ed9f4e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4300},"family":"AA-Briefcase","row_ids":["c6c37db45460e80aac980c7255d8ffed"],"score":55.84},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1400},"family":"AA-AnalystAgent","row_ids":["9cd6a63a2e9021c4431ddd906f3978b5"],"score":60.87}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":86.54,"percentile_exact":{"denominator":13,"numerator":1125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","peer_count":82,"percentile":65.43,"percentile_exact":{"denominator":81,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"4.0","peer_count":76,"percentile":64.0,"percentile_exact":{"denominator":1,"numerator":64},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":78,"percentile":55.84,"percentile_exact":{"denominator":77,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","peer_count":24,"percentile":60.87,"percentile_exact":{"denominator":23,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625}],"score":67.95,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":3822,"numerator":208475},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2600},"family":"FrontierCode","row_ids":["926e007cfcb9b2d7a75e59641d96bd9e"],"score":53.06},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":300},"family":"Vals Code Migration","row_ids":["59c32ec585e3e137a1a5f19c3bda0e98"],"score":23.08},{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":175},"family":"DeepSWE","row_ids":["e589efcb9de8d6dea5758589d780110f"],"score":87.5}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"2.0","peer_count":50,"percentile":53.06,"percentile_exact":{"denominator":49,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":27,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":9,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8}],"score":54.55,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":3835475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1525},"family":"GDP.pdf","row_ids":["1bf434a1cea1439ab199e26911f14113"],"score":80.26},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1450},"family":"SimpleQA Verified","row_ids":["f06072893cbefd23dc63a86432677691"],"score":85.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3100},"family":"AA-Omniscience","row_ids":["1cb384dc9fec91c5db3c1841d018dfc3","4b5e9db335123a5e12d2351e281e6182"],"score":58.49}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916}],"score":74.68,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":38950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2200},"family":"Humanity Sixth Sense","row_ids":["11851315d0cbcd3709f1ea2206b9b74b"],"score":81.48},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1500},"family":"MMMU-Pro","row_ids":["8bcd59ede7a9782bc9b8b3c169c19887"],"score":88.24}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","peer_count":28,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185}],"score":84.86,"weight":10}],"eligible":true,"id":"m-422e5ba87859dd55","in_window":true,"name":"Gemini 3.8 Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":19,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":98.31,"percentile_exact":{"denominator":59,"numerator":5800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","peer_count":39,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","peer_count":17,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","peer_count":52,"percentile":78.43,"percentile_exact":{"denominator":51,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":86.54,"percentile_exact":{"denominator":13,"numerator":1125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","peer_count":82,"percentile":65.43,"percentile_exact":{"denominator":81,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"4.0","peer_count":76,"percentile":64.0,"percentile_exact":{"denominator":1,"numerator":64},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":78,"percentile":55.84,"percentile_exact":{"denominator":77,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","peer_count":24,"percentile":60.87,"percentile_exact":{"denominator":23,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"2.0","peer_count":50,"percentile":53.06,"percentile_exact":{"denominator":49,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":27,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":9,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","peer_count":28,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":68.78,"score_exact":{"denominator":18613913710392,"numerator":1280244857994287},"source_id":"google/gemini-3.8-flash"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":41.67,"evaluator_count":1,"exact":{"denominator":59,"numerator":4750},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4750},"family":"GPQA Diamond","row_ids":["2f5e3b6ebf2112f4f734c9c632e248bf"],"score":80.51}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505}],"score":80.51,"weight":25},{"effective_weight_percent":41.67,"evaluator_count":1,"exact":{"denominator":54,"numerator":3625},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":75},"family":"tau3-Banking","row_ids":["569f96b9f64d6f89c3e721fb138a9646"],"score":75.0},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1600},"family":"GDPval-AA","row_ids":["67407c66c8ccc8ec87ccd4b6ae121aa3"],"score":59.26}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"v2.1","peer_count":82,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33}],"score":67.13,"weight":25},{"effective_weight_percent":16.67,"evaluator_count":1,"exact":{"denominator":17,"numerator":550},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":550},"family":"MMMU-Pro","row_ids":["d7ea5990ff16dd400cf1a637a1392f4c"],"score":32.35}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607}],"score":32.35,"weight":10}],"eligible":true,"id":"m-e060cf24e4176269","in_window":true,"name":"Agnes 2.5 Pro Beta","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":20,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"v2.1","peer_count":82,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"result_updated_at":null,"score":66.91,"score_exact":{"denominator":649944,"numerator":43486475},"source_id":"published/aa:agnes-2-5-pro-beta"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":37128,"numerator":2162375},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":350},"family":"FrontierMath","row_ids":["f2e8485b9221333706b0e9e14cd8c4e7"],"score":38.89},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1975},"family":"Terminal-Bench-Science","row_ids":["ebff4c97ce1c3d30b95d6d855419e5c1"],"score":75.96},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4600},"family":"HLE","row_ids":["2bfa42798a174eb5a88b71f065cd4aef"],"score":73.02},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2300},"family":"CritPt","row_ids":["d9b68c289d550eec511fa758a53ee05a"],"score":45.1}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","peer_count":64,"percentile":73.02,"percentile_exact":{"denominator":63,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","peer_count":52,"percentile":45.1,"percentile_exact":{"denominator":51,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25}],"score":58.24,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114912,"numerator":9263075},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2500},"family":"GDPval-AA","row_ids":["0a034403545d9a6f875f18bdfdb8614b"],"score":92.59},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1375},"family":"Terminal-Bench","row_ids":["6d06cabd5082651a9ea6ff6f7709c136"],"score":57.29},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1650},"family":"AutomationBench","row_ids":["74476994a8c12c8ddef6748c439b2908"],"score":86.84},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":600},"family":"AA-Briefcase","row_ids":["e9014188fe35aeb589dfba302eb5ab58"],"score":85.71}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","peer_count":82,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":78,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42}],"score":80.61,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":49200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3800},"family":"FrontierCode","row_ids":["981d7322c13b53f8d5edfa1fdb6e95e5"],"score":77.55},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1000},"family":"Vals Code Migration","row_ids":["1080a38736b6ba3768e68ace6a340f30"],"score":76.92}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","peer_count":50,"percentile":77.55,"percentile_exact":{"denominator":49,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","peer_count":27,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52}],"score":77.24,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2456375},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1325},"family":"GDP.pdf","row_ids":["b06f9a392143fecd0d26f4f434ebd403"],"score":69.74},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":500},"family":"SimpleQA Verified","row_ids":["8e35d6850446c900d7c667bcebf2d5eb"],"score":29.41},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2350},"family":"AA-Omniscience","row_ids":["28843797b0f403bc00b3b129a0250f86","cd12ad2957c69fc5e8660a959eaae5db"],"score":44.34}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795}],"score":47.83,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":153,"numerator":8575},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"Humanity Sixth Sense","row_ids":["60e8851ddeb919672b146036d9d51182"],"score":44.44},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1150},"family":"MMMU-Pro","row_ids":["33026de8036ee6188f530a53e9aaa0aa"],"score":67.65}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","peer_count":28,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289}],"score":56.05,"weight":10}],"eligible":true,"id":"m-caf1ef1f3bdf5452","in_window":true,"name":"DeepSeek V4.1 Flash 1017","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-17","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":21,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","peer_count":64,"percentile":73.02,"percentile_exact":{"denominator":63,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","peer_count":52,"percentile":45.1,"percentile_exact":{"denominator":51,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","peer_count":82,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":78,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","peer_count":50,"percentile":77.55,"percentile_exact":{"denominator":49,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","peer_count":27,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","peer_count":28,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289}],"release_date":"2026-10-17","release_status":"published","release_verified":true,"result_updated_at":"2026-10-19","score":65.33,"score_exact":{"denominator":37686999168,"numerator":2462117234705},"source_id":"deepseek/deepseek-v4.1-flash-1017"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":152,"numerator":9325},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1500},"family":"Terminal-Bench-Science","row_ids":["f1163968f642895b63614f184b419fdc"],"score":78.95},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":175},"family":"HLE","row_ids":["3ff9ef5159d8b127a6433bdd9ec141a5"],"score":43.75}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","peer_count":39,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8}],"score":61.35,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":1235},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":200},"family":"GDPval-AA","row_ids":["8b095392eb68858c1d89a63205401933"],"score":22.22},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":75},"family":"Terminal-Bench","row_ids":["a142db33093dc7d6f056b8d8509f429e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":40},"family":"AA-Briefcase","row_ids":["ad451031d1d4a9ddfab5c379bfc5d852"],"score":40.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","peer_count":49,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"reference_revision":"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0}],"score":45.74,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["0c33bab4ff416863111d8ddbb4eb7f0b"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8}],"score":100.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1250},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1250},"family":"SimpleQA Verified","row_ids":["8a0eac6d7e9eaac51a5eeee920a35f4d"],"score":73.53}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607}],"score":73.53,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":970},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1400},"family":"Humanity Sixth Sense","row_ids":["e25930e774ad7abbc0f5c7ffcc567aa5"],"score":51.85},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"Chartography","row_ids":["961e2599d1347e8ad57a29581acdb50c"],"score":20.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","peer_count":28,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6}],"score":35.93,"weight":10}],"eligible":true,"id":"m-3b5ea6fdfa94bc08","in_window":true,"name":"GPT-6 Sol","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":22,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","peer_count":39,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","peer_count":49,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"reference_revision":"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","peer_count":28,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":"2026-10-26","score":65.07,"score_exact":{"denominator":93024,"numerator":6053149},"source_id":"openai/gpt-6-sol"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1071,"numerator":66400},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":50},"family":"Terminal-Bench-Science","row_ids":["e374e482dae499c526f6a7da9baef1aa"],"score":50.0},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1600},"family":"HLE","row_ids":["a1e07609863e9a13d3798af4a816cb48"],"score":76.19},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3050},"family":"CritPt","row_ids":["321a79f217d130c0ceb39102c3c1fab9"],"score":59.8}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"0.1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","peer_count":64,"percentile":76.19,"percentile_exact":{"denominator":21,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325}],"score":62.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":474012,"numerator":36496535},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5500},"family":"GDPval-AA","row_ids":["cfa7d9348c146c713f91a2534eae2b29"],"score":67.9},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":70},"family":"Terminal-Bench","row_ids":["3a8fd5af00baca1a19eb582e1a8c4ab5"],"score":70.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1825},"family":"AutomationBench","row_ids":["b04984f7d9488e18c36d59811ea3856b"],"score":96.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5700},"family":"AA-Briefcase","row_ids":["da58b11d089faf8c31c386386035abd7"],"score":74.03}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","peer_count":82,"percentile":67.9,"percentile_exact":{"denominator":81,"numerator":5500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":96.05,"percentile_exact":{"denominator":19,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":78,"percentile":74.03,"percentile_exact":{"denominator":77,"numerator":5700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98}],"score":76.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":2700},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2700},"family":"FrontierCode","row_ids":["1935649843d5e809cef231b07734be32"],"score":55.1}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"2.0","peer_count":50,"percentile":55.1,"percentile_exact":{"denominator":49,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182}],"score":55.1,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":237475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2975},"family":"GDP.pdf","row_ids":["96a35bebfd1cf232725317d8f36f4c53"],"score":78.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["2ab90fc472453676cb1037409906c906","cd0ef2b66ec6818e014beedc4ad88c9f"],"score":39.62}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483}],"score":58.96,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":900},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":900},"family":"MMMU-Pro","row_ids":["fc2b137ca09cf3cd0123e67faaeeed34"],"score":52.94}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786}],"score":52.94,"weight":10}],"eligible":true,"id":"m-8e76831a5814eec9","in_window":true,"name":"MiniMax-M3.1-Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-19","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":23,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"0.1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","peer_count":64,"percentile":76.19,"percentile_exact":{"denominator":21,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","peer_count":82,"percentile":67.9,"percentile_exact":{"denominator":81,"numerator":5500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":96.05,"percentile_exact":{"denominator":19,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":78,"percentile":74.03,"percentile_exact":{"denominator":77,"numerator":5700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"2.0","peer_count":50,"percentile":55.1,"percentile_exact":{"denominator":49,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786}],"release_date":"2026-10-19","release_status":"published","release_verified":true,"result_updated_at":null,"score":62.85,"score_exact":{"denominator":703433808,"numerator":44213635945},"source_id":"minimax/MiniMax-M3.1-Flash"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":3717,"numerator":278450},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5000},"family":"GPQA Diamond","row_ids":["185f07372e1def6a3525e6b1b6c635f5"],"score":84.75},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4100},"family":"HLE","row_ids":["c42c9d1f70b755ec31ff5dcd6c984f87"],"score":65.08}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":84.75,"percentile_exact":{"denominator":59,"numerator":5000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","peer_count":64,"percentile":65.08,"percentile_exact":{"denominator":63,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211}],"score":74.91,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":105856906},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1075},"family":"tau3-Banking","row_ids":["b1886f41be6be37742b893448f0c7326"],"score":82.69},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1900},"family":"GDPval-AA","row_ids":["6a2341b888c3c4634de9d8d25fc69ba9"],"score":70.37},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":172},"family":"Terminal-Bench","row_ids":["cdfb425c1664f272548a144cf79359c8"],"score":57.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1075},"family":"AutomationBench","row_ids":["6ca0b1280bb8cb8f57834c52f2472139"],"score":56.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5200},"family":"AA-Briefcase","row_ids":["94b4404155c32f58241674a265734f95"],"score":67.53},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":700},"family":"EnterpriseOps-Gym","row_ids":["54f26e677fb3a678985d92f416cf28ad"],"score":77.78}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","peer_count":82,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"4.0","peer_count":76,"percentile":57.33,"percentile_exact":{"denominator":3,"numerator":172},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","peer_count":78,"percentile":67.53,"percentile_exact":{"denominator":77,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943}],"score":68.71,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":1050},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1050},"family":"GDP.pdf","row_ids":["b7c40d3349d6abfac4e0b93abf2adfa2"],"score":55.26}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12}],"score":55.26,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":500},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":500},"family":"MMMU-Pro","row_ids":["dbca8fa745f042a216711b700c5e06ca"],"score":29.41}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717}],"score":29.41,"weight":10}],"eligible":true,"id":"m-65378fc44ea2de26","in_window":true,"name":"DeepSeek V4 Flash Vision","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":24,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":84.75,"percentile_exact":{"denominator":59,"numerator":5000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","peer_count":64,"percentile":65.08,"percentile_exact":{"denominator":63,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","peer_count":82,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"4.0","peer_count":76,"percentile":57.33,"percentile_exact":{"denominator":3,"numerator":172},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","peer_count":78,"percentile":67.53,"percentile_exact":{"denominator":77,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717}],"release_date":"2026-08-21","release_status":"published","release_verified":true,"result_updated_at":null,"score":62.38,"score_exact":{"denominator":3090321234,"numerator":192760715305},"source_id":"published/aa:deepseek-v4-flash-vision"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":5967,"numerator":358000},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["6b7d4617c648ad490ad35eb982d8d2e7"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":500},"family":"HLE","row_ids":["39803a6fda08687f289513743c809091"],"score":55.56},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1200},"family":"CritPt","row_ids":["20d35e57ebd245b7509c38232fe59998"],"score":70.59}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","peer_count":64,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","peer_count":52,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385}],"score":60.0,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":158004,"numerator":12977233},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2300},"family":"GDPval-AA","row_ids":["f04f5a2ca29d7ef8a19efd2486d1e259"],"score":85.19},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":224},"family":"Terminal-Bench","row_ids":["cf0c6cf90a55894b4c003eacdea2ccb7"],"score":74.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1675},"family":"AutomationBench","row_ids":["115a839b24d533001f549fec19f24cc9"],"score":88.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6200},"family":"AA-Briefcase","row_ids":["2d7e1c988db8f9f1d1a2378f0202de51"],"score":80.52}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","peer_count":82,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":78,"percentile":80.52,"percentile_exact":{"denominator":77,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522}],"score":82.13,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":2300},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2300},"family":"FrontierCode","row_ids":["a3fd831de30519528e597eb5e2ceea3e"],"score":46.94}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","peer_count":50,"percentile":46.94,"percentile_exact":{"denominator":49,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151}],"score":46.94,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":205325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2225},"family":"GDP.pdf","row_ids":["d7a0ab07d52625e0d345f1bbf275edda"],"score":58.55},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["568a15734298d304532a5f2a4188d632","ae54534276a6c4c19dcb828d96fdee4c"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247}],"score":50.97,"weight":20}],"eligible":true,"id":"m-a3f1f1de993a039e","in_window":true,"name":"Ling 3.1 Flash 1019","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":25,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","peer_count":64,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","peer_count":52,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","peer_count":82,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":78,"percentile":80.52,"percentile_exact":{"denominator":77,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","peer_count":50,"percentile":46.94,"percentile_exact":{"denominator":49,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247}],"release_date":"2026-10-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":61.24,"score_exact":{"denominator":233188307352,"numerator":14280161662615},"source_id":"published/aa:ling-3-1-flash-rc5"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":1064,"numerator":64025},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":400},"family":"FrontierMath","row_ids":["2e3935359cea7e94134e77b2b6c16a83"],"score":57.14},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":75},"family":"GPQA Diamond","row_ids":["06022591029d9808b7b2c118e22277aa"],"score":75.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1350},"family":"Terminal-Bench-Science","row_ids":["5555c117d96b482e6d1bb0b304cbbefd"],"score":71.05},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":75},"family":"HLE","row_ids":["52db35ab0a7c574170e312e2a679c471"],"score":37.5}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v2","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","peer_count":39,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2}],"score":60.17,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":6,"exact":{"denominator":756,"numerator":37175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":500},"family":"APEX-Agents","row_ids":["100759b239352aa8328473780a99c2b5"],"score":71.43},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":400},"family":"GDPval-AA","row_ids":["8aeb7f071b6b514a7531446104ba33a1"],"score":44.44},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":200},"family":"Terminal-Bench","row_ids":["340dbce590b6aaaa8e1dc907ce4c05ef"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":125},"family":"AutomationBench","row_ids":["29e83d6eda99c0b01ee6e4a2561a6131"],"score":62.5},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":250},"family":"AA-Briefcase","row_ids":["02e903393ff93c77cfd706e931424b8b"],"score":27.78},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":200},"family":"MCP Atlas","row_ids":["938965e9ee08fb381a111173123203b2"],"score":22.22}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","included":true,"model_id":"m-0510680735d3c79a","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"2.1","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"4.0","peer_count":49,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v1.1","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8}],"score":49.17,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["ff5c1b3b8cf739706c1637caa41bfd3b"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1.1","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":1,"numerator":70},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"IFBench","row_ids":["3cf0fe91d3392b3852bb737856d256d6"],"score":40.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"GDP.pdf","row_ids":["5f7d8e688eab4e132d7071fdcec04c3f"],"score":100.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7}],"score":70.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":81,"numerator":5600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1100},"family":"Humanity Sixth Sense","row_ids":["8a450f924e67dc0b617d0ac58b1154ec"],"score":40.74},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":100},"family":"MMMU-Pro","row_ids":["267266c10eb3ebd28421d19f9b4fb8da"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"Chartography","row_ids":["c10b5e0559e0dd2ee1b376a9b0908819"],"score":66.67}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","peer_count":28,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9}],"score":69.14,"weight":10}],"eligible":true,"id":"m-0510680735d3c79a","in_window":true,"name":"GPT-5.6 Sol","official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profile_id":"native","rank":26,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v2","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","peer_count":39,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","included":true,"model_id":"m-0510680735d3c79a","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"2.1","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"4.0","peer_count":49,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v1.1","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1.1","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","peer_count":28,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":60.75,"score_exact":{"denominator":344736,"numerator":20942839},"source_id":"openai/gpt-5.6-sol"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":26334,"numerator":1972813},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":700},"family":"GDPval-AA","row_ids":["efd709c75b89ba20347f69709bb9a3a4"],"score":77.78},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":78},"family":"Terminal-Bench","row_ids":["070c8b1a9e5fb50b3ddeb6aba50faf78"],"score":78.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1500},"family":"AutomationBench","row_ids":["712bffe9ecc8daeca7e0dc7a1fe07e58"],"score":78.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5000},"family":"AA-Briefcase","row_ids":["d6fbe929d324cc028c3c012f923590ba"],"score":64.94}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","peer_count":82,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","peer_count":76,"percentile":78.0,"percentile_exact":{"denominator":1,"numerator":78},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","peer_count":78,"percentile":64.94,"percentile_exact":{"denominator":77,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94}],"score":74.92,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":38,"numerator":1625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1625},"family":"GDP.pdf","row_ids":["d7e50c008783e8a4b24ebddf0211490f"],"score":42.76}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"score":42.76,"weight":20}],"eligible":true,"id":"m-f74476b1532b3a4f","in_window":true,"name":"Ling 3.1 Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":27,"records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","peer_count":82,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","peer_count":76,"percentile":78.0,"percentile_exact":{"denominator":1,"numerator":78},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","peer_count":78,"percentile":64.94,"percentile_exact":{"denominator":77,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"release_date":"2026-10-01","release_status":"published","release_verified":true,"result_updated_at":null,"score":60.63,"score_exact":{"denominator":237006,"numerator":14368565},"source_id":"inclusionai/ling-3.1-flash"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":1200591,"numerator":76127425},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5100},"family":"GPQA Diamond","row_ids":["85f2af62bad122fb48d79f2a5cd0ba76"],"score":86.44},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["6cbde617319ab4da7917bf56251c3a2e"],"score":44.74},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3700},"family":"HLE","row_ids":["20084307dcc6cac0bc9c76758b5e2139"],"score":58.73},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3250},"family":"CritPt","row_ids":["6177f516623a8dfc4d1c35316f9c20f2"],"score":63.73}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":86.44,"percentile_exact":{"denominator":59,"numerator":5100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","peer_count":64,"percentile":58.73,"percentile_exact":{"denominator":63,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335}],"score":63.41,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":4,"exact":{"denominator":196240968,"numerator":14270932625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":250},"family":"APEX-Agents","row_ids":["21a8aa59fcb39de6e961594401b8acf8"],"score":83.33},{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":100},"family":"tau3-Banking","row_ids":["0034bad285685afff6ae0af098ac3976"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7000},"family":"GDPval-AA","row_ids":["8faa2a90e5a7df630665e90e7dc8da8c"],"score":86.42},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1675},"family":"Terminal-Bench","row_ids":["379465a1dce0780ec05eea7fd360c800"],"score":69.79},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1525},"family":"AutomationBench","row_ids":["a4b22debd5eda444528d12c717ed41a1"],"score":80.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6000},"family":"AA-Briefcase","row_ids":["98953845b6332ba7948a61f814a750d3"],"score":77.92},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["4a39f12346421fa0779ea6e76813ff8f"],"score":33.33},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":700},"family":"MCP Atlas","row_ids":["442d36ffd8ae4e2ebc285d09f51efdf2"],"score":77.78},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1050},"family":"AA-AnalystAgent","row_ids":["18b8a0177c3660f9aaf285f2b3571603"],"score":45.65}],"family_count":9,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","peer_count":82,"percentile":86.42,"percentile_exact":{"denominator":81,"numerator":7000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":78,"percentile":77.92,"percentile_exact":{"denominator":77,"numerator":6000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"score":72.72,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":273,"numerator":16000},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":400},"family":"FrontierCode","row_ids":["326d68113ecd602bd7d85d3d2b454d3f"],"score":57.14},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":800},"family":"Vals Code Migration","row_ids":["78dbe912ae9dbb4dfacad05113fef49e"],"score":61.54},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":400},"family":"DeepSWE","row_ids":["a1f523e170da53cbcf5ad3cbcb817e6a"],"score":57.14}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"2.0","peer_count":50,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1.1","peer_count":27,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","included":true,"model_id":"m-c65e3e17492500e9","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1}],"score":58.61,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":34238,"numerator":1354975},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["ba325e0068d64ba19ce176fe95dbb3e4"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":400},"family":"SimpleQA Verified","row_ids":["dd5c55d1f80c2ba62c676b7e3e150a64"],"score":23.53},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2500},"family":"AA-Omniscience","row_ids":["be51199fdc7d1731045d427201852660","330a13ca3aada5b9674ba75668e0acfc"],"score":47.17}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838}],"score":39.58,"weight":20}],"eligible":true,"id":"m-c65e3e17492500e9","in_window":true,"name":"GLM-5.3","official_evidence":{"artifact_sha256":"2dd162815ce7015e44d62c786bf3b76ffde8b7dc7834b1cd6a791f78c3d68667","basis":"official_release_page","date":"2026-08-14","sha256":"2dd162815ce7015e44d62c786bf3b76ffde8b7dc7834b1cd6a791f78c3d68667","url":"https://www.zhipuai.cn/zh/research/162"},"primary_profile_id":"native","profile_id":"native","rank":28,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":86.44,"percentile_exact":{"denominator":59,"numerator":5100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","peer_count":64,"percentile":58.73,"percentile_exact":{"denominator":63,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","peer_count":82,"percentile":86.42,"percentile_exact":{"denominator":81,"numerator":7000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":78,"percentile":77.92,"percentile_exact":{"denominator":77,"numerator":6000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"2.0","peer_count":50,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1.1","peer_count":27,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","included":true,"model_id":"m-c65e3e17492500e9","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838}],"release_date":"2026-08-14","release_status":"published","release_verified":true,"result_updated_at":"2026-08-28","score":59.63,"score_exact":{"denominator":2441081826591408,"numerator":145567376505820175},"source_id":"zhipuai/glm-5.3"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":2570275},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":350},"family":"FrontierMath","row_ids":["21e96013106dd82d62904afe0c4f768b"],"score":38.89},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["0e248a56228a0b9ac7b762f2f3b00fc9"],"score":57.89},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5000},"family":"HLE","row_ids":["de9339207cccf727d5f73de793318c18"],"score":79.37},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1300},"family":"CritPt","row_ids":["b0085122202e14e07e4a784b0954985f"],"score":76.47}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Diamond","peer_count":64,"percentile":79.37,"percentile_exact":{"denominator":63,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","peer_count":52,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435}],"score":63.15,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":143451,"numerator":9916043},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2200},"family":"GDPval-AA","row_ids":["ac3f66c4919122d1f31985187718ed07"],"score":81.48},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":224},"family":"Terminal-Bench","row_ids":["7706b3202bdd1e09b6c5e4786d4bb3e4"],"score":74.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"AutomationBench","row_ids":["9f3e53e10beed094c8835d4e672cc44f"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6800},"family":"AA-Briefcase","row_ids":["6fc39f46271fccbf5e9ec9022c726f0f"],"score":88.31},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":100},"family":"OSWorld","row_ids":["5b785287811a33de601a7060b80bab3e"],"score":33.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2000},"family":"AA-AnalystAgent","row_ids":["a7b164121b0a9a4769bf158d9212d770"],"score":86.96}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2.1","peer_count":82,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":78,"percentile":88.31,"percentile_exact":{"denominator":77,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.1","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","peer_count":24,"percentile":86.96,"percentile_exact":{"denominator":23,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375}],"score":69.12,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":91,"numerator":6350},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":600},"family":"FrontierCode","row_ids":["cbf1fa28fdc821e9836816e2dbc59c73"],"score":85.71},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":700},"family":"Vals Code Migration","row_ids":["c1b5961685ee27b7ccd1b99becf1d743"],"score":53.85}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.0","peer_count":50,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"1.1","peer_count":27,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36}],"score":69.78,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":4819875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2975},"family":"GDP.pdf","row_ids":["a183b88682c41fc78cc2ad5e5485bbb7"],"score":78.29},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":100},"family":"SimpleQA Verified","row_ids":["a7e4020697adbd00e97487d7c55696e0"],"score":5.88},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"MultiChallenge","row_ids":["5d9b045cf9ce01ea061077e1b2e84c1e"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3000},"family":"AA-Omniscience","row_ids":["30bc019f7dc3e7314870de193c23616d","b11d60a588bf725879871b88bc8f70cb"],"score":56.6}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897}],"score":35.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"Chartography","row_ids":["888dd585273ea6255f0bebeb11182076"],"score":50.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4}],"score":50.0,"weight":10}],"eligible":true,"id":"m-016da3c7828e329e","in_window":true,"name":"Claude Haiku 5.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-07","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":29,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Diamond","peer_count":64,"percentile":79.37,"percentile_exact":{"denominator":63,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","peer_count":52,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2.1","peer_count":82,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":78,"percentile":88.31,"percentile_exact":{"denominator":77,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.1","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","peer_count":24,"percentile":86.96,"percentile_exact":{"denominator":23,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.0","peer_count":50,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"1.1","peer_count":27,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4}],"release_date":"2026-10-07","release_status":"published","release_verified":true,"result_updated_at":"2026-10-09","score":59.06,"score_exact":{"denominator":127698358788,"numerator":7542473801321},"source_id":"anthropic/claude-haiku-5-5"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":50},"family":"Terminal-Bench-Science","row_ids":["174d2abbe0e35fdfac2a047ae2fa2208"],"score":50.0}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","peer_count":39,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14}],"score":50.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":72,"numerator":2375},"families":[{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":325},"family":"Terminal-Bench","row_ids":["d7c99e26bef72e8691defabf85dff136"],"score":27.08},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":350},"family":"AA-Briefcase","row_ids":["d43d801b58cb460394b97c15079fc130"],"score":38.89}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","peer_count":49,"percentile":27.08,"percentile_exact":{"denominator":12,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365}],"score":32.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"run 0.1.200","exact":{"denominator":2,"numerator":125},"family":"MirrorCode","row_ids":["02fcbee985f3da4b53d62ce022f2a290"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1400},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1400},"family":"SimpleQA Verified","row_ids":["ab75c5704ac20fa5dc1c93828558d2ce"],"score":82.35}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692}],"score":82.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":27,"numerator":2000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2000},"family":"Humanity Sixth Sense","row_ids":["e0d9bd2a6ca55a94058a8fdfafb19d29"],"score":74.07}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","peer_count":28,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8}],"score":74.07,"weight":10}],"eligible":true,"id":"m-16a8dce2a6e7167a","in_window":true,"name":"Gemini 3.7 Flash","official_evidence":{"artifact_sha256":"b68cd4829f86d5c5a5cafdbccf2e60b6956a5bfc04074d538c1dd7d5c4a9a559","basis":"official_release_page","date":"2026-08-13","sha256":"b68cd4829f86d5c5a5cafdbccf2e60b6956a5bfc04074d538c1dd7d5c4a9a559","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash/"},"primary_profile_id":"native","profile_id":"native","rank":30,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","peer_count":39,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","peer_count":49,"percentile":27.08,"percentile_exact":{"denominator":12,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","peer_count":28,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8}],"release_date":"2026-08-13","release_status":"published","release_verified":true,"result_updated_at":"2026-10-21","score":57.12,"score_exact":{"denominator":14688,"numerator":839045},"source_id":"google/gemini-3.7-flash"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":4200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4200},"family":"GPQA Diamond","row_ids":["004e3b6a13ca6828e88e6f4e76a97e76"],"score":71.19}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":71.19,"percentile_exact":{"denominator":59,"numerator":4200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192}],"score":71.19,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3081078,"numerator":117236125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["0456438c8b02eed2e908e4593e41dcf0"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3700},"family":"GDPval-AA","row_ids":["9050cb18d0244c05aa4a3bdecef069eb"],"score":45.68},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":700},"family":"AutomationBench","row_ids":["c9e2b1db45e7b21cbd9a431702d99eb1"],"score":36.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2700},"family":"AA-Briefcase","row_ids":["82a245d0d5111b9661f660b2a94b3db6"],"score":35.06}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v2.1","peer_count":82,"percentile":45.68,"percentile_exact":{"denominator":81,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v1.1","peer_count":78,"percentile":35.06,"percentile_exact":{"denominator":77,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26}],"score":38.05,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":125},"family":"IFBench","row_ids":["353142d52584e4a6aac9d52541e06d23"],"score":62.5},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2225},"family":"GDP.pdf","row_ids":["312a7fa68250f3fcf7a32404595e6dc4"],"score":58.55}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126}],"score":60.53,"weight":20}],"eligible":true,"id":"m-c0755395b2da9721","in_window":true,"name":"Nex-N2-Pro","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":31,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":71.19,"percentile_exact":{"denominator":59,"numerator":4200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v2.1","peer_count":82,"percentile":45.68,"percentile_exact":{"denominator":81,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v1.1","peer_count":78,"percentile":35.06,"percentile_exact":{"denominator":77,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126}],"release_date":"2026-06-02","release_status":"published","release_verified":true,"result_updated_at":null,"score":56.31,"score_exact":{"denominator":2544970428,"numerator":143298061675},"source_id":"nex-agi/nex-n2-pro"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":7075},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":175},"family":"GPQA Diamond","row_ids":["7ff900665f3a2edd6bd47eeb54dcc81e"],"score":58.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1250},"family":"Terminal-Bench-Science","row_ids":["a18d9b692129832a6701c8778a0404f3"],"score":65.79}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":58.33,"percentile_exact":{"denominator":3,"numerator":175},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","peer_count":39,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86}],"score":62.06,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":1008,"numerator":58825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":200},"family":"APEX-Agents","row_ids":["bf4bf643f4907b71184823fd0ebd6b1c"],"score":28.57},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":725},"family":"Terminal-Bench","row_ids":["e085a37b9093e9f4b5222ae848bc472f"],"score":60.42},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"AutomationBench","row_ids":["58a53dd6f007e225a58569d30c64eee8"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":850},"family":"AA-Briefcase","row_ids":["0d1d30147df15075a1ed15467a0ef1b4"],"score":94.44}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","included":true,"model_id":"m-99ca714b8835cfb9","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","peer_count":49,"percentile":60.42,"percentile_exact":{"denominator":12,"numerator":725},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"reference_revision":"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593}],"score":58.36,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["9619717553c9af1d9fdaa3ac01eac67f"],"score":25.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"reference_revision":"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6}],"score":25.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":17,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["57da33d0598ebe148085f216c7c7cefb"],"score":100.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":600},"family":"SimpleQA Verified","row_ids":["2e690ecc6ebeb390f087e351bd8c7803"],"score":35.29}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458}],"score":67.65,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":54,"numerator":3725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1700},"family":"Humanity Sixth Sense","row_ids":["b0de0689b04f1b90ca42e94e49e0a3bb"],"score":62.96},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":75},"family":"MMMU-Pro","row_ids":["d1f2e18c925df69cd229259da568a231"],"score":75.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","peer_count":28,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3}],"score":68.98,"weight":10}],"eligible":true,"id":"m-99ca714b8835cfb9","in_window":true,"name":"Qwen3.8 Max","official_evidence":{"artifact_sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","basis":"official_release_page","date":"2026-08-02","sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","url":"https://www.alibabacloud.com/help/en/model-studio/newly-released-models"},"primary_profile_id":"native","profile_id":"native","rank":32,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":58.33,"percentile_exact":{"denominator":3,"numerator":175},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","peer_count":39,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","included":true,"model_id":"m-99ca714b8835cfb9","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","peer_count":49,"percentile":60.42,"percentile_exact":{"denominator":12,"numerator":725},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"reference_revision":"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"reference_revision":"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","peer_count":28,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3}],"release_date":"2026-08-02","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":55.53,"score_exact":{"denominator":3907008,"numerator":216965705},"source_id":"alibaba/qwen3.8-max"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":19,"numerator":1200},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1200},"family":"Terminal-Bench-Science","row_ids":["545c704b3f9b5b5c3a00fbe6211a666c"],"score":63.16}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","peer_count":39,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43}],"score":63.16,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":144,"numerator":6875},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":275},"family":"Terminal-Bench","row_ids":["620e1ee85f651d96e2f6e21d676ecc23"],"score":34.38},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":550},"family":"AA-Briefcase","row_ids":["45f7b80c7cc89c267cb8e0d74593b6e4"],"score":61.11}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202}],"score":47.74,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"run 0.1.200","exact":{"denominator":2,"numerator":125},"family":"MirrorCode","row_ids":["0621e61b8874eac3232c0422d1f71677"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111}],"score":62.5,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":17,"numerator":800},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":800},"family":"SimpleQA Verified","row_ids":["954b3bb3d889d2d8fa0f84b9a3223eaf"],"score":47.06}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493}],"score":47.06,"weight":20}],"eligible":true,"id":"m-3c17c0d3ca7f1abe","in_window":true,"name":"Grok 4.6","official_evidence":{"artifact_sha256":"22a9c3dcee05d0c5a93b52c78e1c1e867278c4f39ebda057faec281332c01411","basis":"official_release_page","date":"2026-08-12","sha256":"22a9c3dcee05d0c5a93b52c78e1c1e867278c4f39ebda057faec281332c01411","url":"https://x.ai/news/grok-4-6"},"primary_profile_id":"native","profile_id":"native","rank":33,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","peer_count":39,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":"2026-10-21","score":55.15,"score_exact":{"denominator":837216,"numerator":46174325},"source_id":"xai/grok-4.6"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1642914,"numerator":100299025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4900},"family":"GPQA Diamond","row_ids":["3551c30b97ba4cb5f0fc94e087d94b59"],"score":83.05},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":750},"family":"Terminal-Bench-Science","row_ids":["0ca35771477057319be8d71ddeaad930"],"score":57.69},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3800},"family":"HLE","row_ids":["a57cb2704c93293d339e91362f40ae46"],"score":60.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2200},"family":"CritPt","row_ids":["620b6f3e6314727d16068f5503869955"],"score":43.14}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":83.05,"percentile_exact":{"denominator":59,"numerator":4900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","peer_count":64,"percentile":60.32,"percentile_exact":{"denominator":63,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","peer_count":52,"percentile":43.14,"percentile_exact":{"denominator":51,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245}],"score":61.05,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":141729588,"numerator":8766837125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":2375},"family":"tau3-Banking","row_ids":["9a44264bd8b89c341b6a8bd9496be3bf"],"score":91.35},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6800},"family":"GDPval-AA","row_ids":["3091d538e8d1e5edf6eea60314cea6a6"],"score":83.95},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":775},"family":"Terminal-Bench","row_ids":["28abab0f2a19e622cf15ffcdc9be3dfd"],"score":64.58},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1475},"family":"AutomationBench","row_ids":["e492bf3e10afd1293a1152cd005f96a3"],"score":77.63},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":800},"family":"AA-Briefcase","row_ids":["3015c7ad3cd2d55a621f41dcebd98f3d"],"score":72.73},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["b8b704019523524af0ce859e89927fa0"],"score":16.67},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":600},"family":"AA-AnalystAgent","row_ids":["a9057fe2d742aeb1aa0666773aa5958d"],"score":26.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","peer_count":82,"percentile":83.95,"percentile_exact":{"denominator":81,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","peer_count":49,"percentile":64.58,"percentile_exact":{"denominator":12,"numerator":775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":77.63,"percentile_exact":{"denominator":19,"numerator":1475},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":78,"percentile":72.73,"percentile_exact":{"denominator":11,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","peer_count":24,"percentile":26.09,"percentile_exact":{"denominator":23,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25}],"score":61.86,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":30950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2500},"family":"FrontierCode","row_ids":["32536ff0b3b8576673f9f05323026239"],"score":51.02},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":600},"family":"Vals Code Migration","row_ids":["efd99741d3524e2d5ab3b679f8da8006"],"score":46.15}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","peer_count":50,"percentile":51.02,"percentile_exact":{"denominator":49,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","peer_count":27,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65}],"score":48.59,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":108625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1225},"family":"GDP.pdf","row_ids":["f95dbe30b012b632cab81f4598931983"],"score":64.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["a76a9fe9ce50e17717765fcaebd6fbf8","7e86ee60903a41ed29953982d04673a1"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499}],"score":53.93,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["7d2a87235efadfc965d04473b687dd29"],"score":33.33}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","peer_count":28,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8}],"score":33.33,"weight":10}],"eligible":true,"id":"m-cf177eb625d4bd3e","in_window":true,"name":"GLM-5.3-Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":34,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":83.05,"percentile_exact":{"denominator":59,"numerator":4900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","peer_count":64,"percentile":60.32,"percentile_exact":{"denominator":63,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","peer_count":52,"percentile":43.14,"percentile_exact":{"denominator":51,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","peer_count":82,"percentile":83.95,"percentile_exact":{"denominator":81,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","peer_count":49,"percentile":64.58,"percentile_exact":{"denominator":12,"numerator":775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":77.63,"percentile_exact":{"denominator":19,"numerator":1475},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":78,"percentile":72.73,"percentile_exact":{"denominator":11,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","peer_count":24,"percentile":26.09,"percentile_exact":{"denominator":23,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","peer_count":50,"percentile":51.02,"percentile_exact":{"denominator":49,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","peer_count":27,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","peer_count":28,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"result_updated_at":null,"score":54.56,"score_exact":{"denominator":210957688717776,"numerator":11510724316142215},"source_id":"zhipuai/glm-5.3-flash"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":399,"numerator":23825},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":600},"family":"FrontierMath","row_ids":["51cb1dea03ece9a899f843aedba7597e"],"score":85.71},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":25},"family":"GPQA Diamond","row_ids":["b6943c0f15a31bf97046c2e4af0b90f3"],"score":25.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1300},"family":"Terminal-Bench-Science","row_ids":["f5cc9d3986b503f55588d97bf5fde40a"],"score":68.42}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","peer_count":8,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","peer_count":39,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71}],"score":59.71,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":144,"numerator":8225},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["b82b11faf1fed2434465112eb548316c"],"score":100.0},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":875},"family":"Terminal-Bench","row_ids":["5ecd3796946e727694a868a88027022a"],"score":72.92},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["8ad20dc033b0a09d811fcebfca51c2ab"],"score":0.0},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":500},"family":"MCP Atlas","row_ids":["37801ada9ba8d94f8b7e8f5a6667ee11"],"score":55.56}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","peer_count":49,"percentile":72.92,"percentile_exact":{"denominator":12,"numerator":875},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3}],"score":57.12,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"DeepSWE","row_ids":["0772cd4ebc7a33c01cf749b6a69bd7c0"],"score":50.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","peer_count":9,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9}],"score":50.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":2,"numerator":95},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"IFBench","row_ids":["3f5e26906c244a4d4f9e57725da3d9df"],"score":20.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"GDP.pdf","row_ids":["134a991b936b3bccddf2a9a60c450e69"],"score":75.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8}],"score":47.5,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":3,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["cf924b68bd130c394ca909ce8970cb90"],"score":66.67},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["a6ccc017ec5976d43545f1578c4ba3f8"],"score":50.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","peer_count":28,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2}],"score":58.33,"weight":10}],"eligible":true,"id":"m-86c0c8bf10aa4a0d","in_window":true,"name":"Claude Fable 5","official_evidence":{"artifact_sha256":"c8267145b47f034cb45aedbe8074a7b244fee4ab2050099a2f6e2e7f9d6d9ed5","basis":"official_release_page","date":"2026-06-09","sha256":"c8267145b47f034cb45aedbe8074a7b244fee4ab2050099a2f6e2e7f9d6d9ed5","url":"https://www.anthropic.com/news/claude-fable-5-mythos-5"},"primary_profile_id":"native","profile_id":"native","rank":35,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","peer_count":8,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","peer_count":39,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","peer_count":49,"percentile":72.92,"percentile_exact":{"denominator":12,"numerator":875},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","peer_count":9,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","peer_count":28,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2}],"release_date":"2026-06-09","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":54.54,"score_exact":{"denominator":76608,"numerator":4178261},"source_id":"anthropic/claude-fable-5"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["f0716aa67008d67c9923512a4d8c5e11"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":162,"numerator":3275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["01aeec9c08a057635f9b72825df4f3fb"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":50},"family":"tau3-Banking","row_ids":["e24b9f9c121b7281ff7f37b4b404f354"],"score":50.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2500},"family":"GDPval-AA","row_ids":["7bf77397fcea6af7a97a2524f4c194bb"],"score":30.86},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["b96a70ae7b2ded290b5f169ceedf1e03"],"score":0.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","peer_count":82,"percentile":30.86,"percentile_exact":{"denominator":81,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2}],"score":20.22,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":9,"numerator":850},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"IFBench","row_ids":["dafb6731eb5f8725711540078af7013b"],"score":88.89},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"MMLU-Pro","row_ids":["27d4cca7113cd0216eb12449ea9867c2"],"score":100.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2}],"score":94.44,"weight":20}],"eligible":true,"id":"m-b7faaf31060e95f7","in_window":true,"name":"Solar Open2 250B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":36,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","peer_count":82,"percentile":30.86,"percentile_exact":{"denominator":81,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":54.18,"score_exact":{"denominator":133812,"numerator":7249925},"source_id":"published/aa:solar-open2-250b"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":20349,"numerator":1015100},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":350},"family":"Terminal-Bench-Science","row_ids":["f80ea7a9c8b1ad0f866390a070dec6ae"],"score":18.42},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":500},"family":"HLE","row_ids":["ea28428a53ddb22639d7f5b7e50547a2"],"score":71.43},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3050},"family":"CritPt","row_ids":["b5e83ad92c83f666dbcd3cdc392d63b6"],"score":59.8}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","included":true,"model_id":"m-f892623396612d9b","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","included":true,"model_id":"m-f892623396612d9b","original_version_label":"Diamond","peer_count":64,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325}],"score":49.88,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":152631864,"numerator":9897530725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"APEX-Agents","row_ids":["519c33c19b165c6197391af117e522e3"],"score":75.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5900},"family":"GDPval-AA","row_ids":["d52704f212d466a4d653d53f1794d1db"],"score":72.84},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1375},"family":"Terminal-Bench","row_ids":["d5a9a67230f3e72676fe2415a004dd47"],"score":57.29},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1175},"family":"AutomationBench","row_ids":["85e95115c24e6ebd5d7a72360c176406"],"score":61.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5300},"family":"AA-Briefcase","row_ids":["5342274720726a814f5f5d822f07b510"],"score":68.83},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":250},"family":"EnterpriseOps-Gym","row_ids":["6be4e7acd5ffa2c31291fe47e51a2a60"],"score":83.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":800},"family":"AA-AnalystAgent","row_ids":["1f372ec6dfb68a53b6ae42bbb816dddf"],"score":34.78}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2.1","peer_count":82,"percentile":72.84,"percentile_exact":{"denominator":81,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","included":true,"model_id":"m-f892623396612d9b","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","included":true,"model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":78,"percentile":68.83,"percentile_exact":{"denominator":77,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","peer_count":24,"percentile":34.78,"percentile_exact":{"denominator":23,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"score":64.85,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":25250},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2000},"family":"FrontierCode","row_ids":["4086551829a96ee8ea164e44d1e7c1c5"],"score":40.82},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":500},"family":"Vals Code Migration","row_ids":["68fce01452fe549878e8933d04edaa05"],"score":38.46}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","included":true,"model_id":"m-f892623396612d9b","original_version_label":"2.0","peer_count":50,"percentile":40.82,"percentile_exact":{"denominator":49,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","included":true,"model_id":"m-f892623396612d9b","original_version_label":"1.1","peer_count":27,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0}],"score":39.64,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":62200},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1200},"family":"GDP.pdf","row_ids":["32f9ed0b80b3e83e515a9f3fb34bc9b3"],"score":63.16},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3200},"family":"AA-Omniscience","row_ids":["ec9021441721b216b9626f97b96062f1","880d1169acb398e66460aed2eb721de0"],"score":60.38}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072}],"score":61.77,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":650},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":650},"family":"MMMU-Pro","row_ids":["e0d9cab1ce675bebaf2c47a0db9ddea9"],"score":38.24}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902}],"score":38.24,"weight":10}],"eligible":true,"id":"m-f892623396612d9b","in_window":true,"name":"Step 5 Preview","official_evidence":{"artifact_sha256":"f316d16c93b2d4ce98e733917634947e32a9218037b9e1d3d427de4b75275ddb","basis":"official_release_page","date":"2026-09-20","sha256":"f316d16c93b2d4ce98e733917634947e32a9218037b9e1d3d427de4b75275ddb","url":"https://www.stepfun.com/news/step-5-preview"},"primary_profile_id":"native","profile_id":"native","rank":37,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","included":true,"model_id":"m-f892623396612d9b","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","included":true,"model_id":"m-f892623396612d9b","original_version_label":"Diamond","peer_count":64,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2.1","peer_count":82,"percentile":72.84,"percentile_exact":{"denominator":81,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","included":true,"model_id":"m-f892623396612d9b","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","included":true,"model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":78,"percentile":68.83,"percentile_exact":{"denominator":77,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","peer_count":24,"percentile":34.78,"percentile_exact":{"denominator":23,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","included":true,"model_id":"m-f892623396612d9b","original_version_label":"2.0","peer_count":50,"percentile":40.82,"percentile_exact":{"denominator":49,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","included":true,"model_id":"m-f892623396612d9b","original_version_label":"1.1","peer_count":27,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902}],"release_date":"2026-09-20","release_status":"published","release_verified":true,"result_updated_at":null,"score":52.79,"score_exact":{"denominator":376374110112,"numerator":19867815404455},"source_id":"stepfun/step-5-preview"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":3591,"numerator":160750},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":350},"family":"Terminal-Bench-Science","row_ids":["17afba9dea11c1bfb4fcda8f7ceb8fe4"],"score":18.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3100},"family":"HLE","row_ids":["2e68084b0515fdc6a19fe2b7a59814d2"],"score":49.21},{"benchmark_version":"2 Hard","exact":{"denominator":3,"numerator":200},"family":"CritPt","row_ids":["2f0eebfa718074ef0cab557ff07d209b"],"score":66.67}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","peer_count":64,"percentile":49.21,"percentile_exact":{"denominator":63,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","peer_count":52,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34}],"score":44.76,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":605682,"numerator":40432325},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":800},"family":"GDPval-AA","row_ids":["82bb5fcac208ae67d3e4a7d3de124d5d"],"score":88.89},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":325},"family":"Terminal-Bench","row_ids":["65da79b527ae8b650ca14c9d0ac7489d"],"score":54.17},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1375},"family":"AutomationBench","row_ids":["45f30ef9c9577f4abce574fa0b12543a"],"score":72.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6100},"family":"AA-Briefcase","row_ids":["c6fb2c23c770d5e7bda4e380c2517fe7"],"score":79.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":900},"family":"AA-AnalystAgent","row_ids":["59ce77c04181c398c36c852841fb4ecf"],"score":39.13}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","peer_count":82,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","peer_count":49,"percentile":54.17,"percentile_exact":{"denominator":6,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":78,"percentile":79.22,"percentile_exact":{"denominator":77,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","peer_count":24,"percentile":39.13,"percentile_exact":{"denominator":23,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375}],"score":66.76,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":34275},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3200},"family":"FrontierCode","row_ids":["f7aceff949a2ebec04ba4fe1d1d7e419"],"score":65.31},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":550},"family":"Vals Code Migration","row_ids":["c811b7493c2ca32eedcc762a5f354d58"],"score":42.31}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","peer_count":50,"percentile":65.31,"percentile_exact":{"denominator":49,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","peer_count":27,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79}],"score":53.81,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":58000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1400},"family":"GDP.pdf","row_ids":["0c01c6860fffc32ef81ab819ece21423"],"score":73.68},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2200},"family":"AA-Omniscience","row_ids":["d99f9179103c8156a9cedc7180b05c42","b3118336d86ca34d1fd6619d5a595882"],"score":41.51}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376}],"score":57.6,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":200},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["73994f722364016b18f62a10e6695b38"],"score":22.22}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","peer_count":28,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2}],"score":22.22,"weight":10}],"eligible":true,"id":"m-dabee5e3d87d8a7a","in_window":true,"name":"MiMo-V2.6-Pro","official_evidence":{"artifact_sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","basis":"official_release_page","date":"2026-09-22","sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","url":"https://mimo.mi.com/docs/en-US/updates/model"},"primary_profile_id":"native","profile_id":"native","rank":38,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","peer_count":64,"percentile":49.21,"percentile_exact":{"denominator":63,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","peer_count":52,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","peer_count":82,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","peer_count":49,"percentile":54.17,"percentile_exact":{"denominator":6,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":78,"percentile":79.22,"percentile_exact":{"denominator":77,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","peer_count":24,"percentile":39.13,"percentile_exact":{"denominator":23,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","peer_count":50,"percentile":65.31,"percentile_exact":{"denominator":49,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","peer_count":27,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","peer_count":28,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":52.38,"score_exact":{"denominator":35054451432,"numerator":1836253825765},"source_id":"xiaomi/mimo-v2.6-pro"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3717,"numerator":175250},"families":[{"benchmark_version":"2026","exact":{"denominator":1,"numerator":100},"family":"AIME","row_ids":["0f70d96aa863b344234e9f6fb914e9f2"],"score":100.0},{"benchmark_version":"500-item subset","exact":{"denominator":3,"numerator":200},"family":"MATH-500","row_ids":["1f0d5271a4aa3a380367a096ae7ae210"],"score":66.67},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1200},"family":"GPQA Diamond","row_ids":["d95e6a6f385be987c8949ca6952efa66"],"score":20.34},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":100},"family":"HLE","row_ids":["75182b4b31b9383c53f3f18d0e09c65d"],"score":1.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"2026","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":20.34,"percentile_exact":{"denominator":59,"numerator":1200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond","peer_count":64,"percentile":1.59,"percentile_exact":{"denominator":63,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038}],"score":47.15,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":18486468,"numerator":627714865},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":1225},"family":"tau3-Banking","row_ids":["dc8c6983f496e505270c3f1d2e08a1c1"],"score":47.12},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1600},"family":"GDPval-AA","row_ids":["f2ca1c75e7e337b37337e6b5eb3f5cd5"],"score":19.75},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["49929d3422a54a6833b2d573268b4f8f"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":225},"family":"AutomationBench","row_ids":["4a74a50d45014989b724a431c32f6c56"],"score":11.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":900},"family":"AA-Briefcase","row_ids":["8a20412103b559527bf3c1a6c26511e5"],"score":11.69},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["9fed535a16c1cf8e07c033c97e3a6783"],"score":100.0}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","included":true,"model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v2.1","peer_count":82,"percentile":19.75,"percentile_exact":{"denominator":81,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","included":true,"model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v1.1","peer_count":78,"percentile":11.69,"percentile_exact":{"denominator":77,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v4","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6}],"score":33.96,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"release_v6","exact":{"denominator":1,"numerator":100},"family":"LiveCodeBench","row_ids":["97029c45627dc4bbce9101042fef65e2"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1}],"score":100.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":171,"numerator":5495},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"IFBench","row_ids":["30e5129b448d30519f98acdcb4661feb"],"score":44.44},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":100},"family":"Multi-IF","row_ids":["ca424296b6739f797da253012aa25cc9"],"score":33.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":150},"family":"GDP.pdf","row_ids":["9ffb2abc9760a5a8e507be89d245bb8f"],"score":7.89},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"IFEval","row_ids":["c5c1fcc77312c8bfdb8f9a2733828c50"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":75},"family":"MMLU-Pro","row_ids":["fdfed7f2ef133f7ec35f427f9ee8f813"],"score":75.0}],"family_count":5,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8}],"score":32.13,"weight":20}],"eligible":true,"id":"m-407d9206347b790d","in_window":true,"name":"MiniCPM5-2B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-07","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":39,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"2026","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":20.34,"percentile_exact":{"denominator":59,"numerator":1200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond","peer_count":64,"percentile":1.59,"percentile_exact":{"denominator":63,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","included":true,"model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v2.1","peer_count":82,"percentile":19.75,"percentile_exact":{"denominator":81,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","included":true,"model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v1.1","peer_count":78,"percentile":11.69,"percentile_exact":{"denominator":77,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v4","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8}],"release_date":"2026-09-07","release_status":"published","release_verified":true,"result_updated_at":null,"score":51.89,"score_exact":{"denominator":19632629016,"numerator":1018776441535},"source_id":"openbmb/minicpm5-2b"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":6571656,"numerator":314311525},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4750},"family":"GPQA Diamond","row_ids":["ba847bf9e90b2e6c2d3a1b021f6f1ff2"],"score":80.51},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["2f71320c4fe6daaec73fc9a09a2e75a6"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3200},"family":"HLE","row_ids":["ee3f7405021072ff698c63b6e1fa510f"],"score":50.79},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":350},"family":"CritPt","row_ids":["a3193e796745f7cc574aae69daceb5f1"],"score":20.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","included":true,"model_id":"m-c947385433db3f03","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond","peer_count":64,"percentile":50.79,"percentile_exact":{"denominator":63,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04}],"score":47.83,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":220077,"numerator":14695565},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1250},"family":"tau3-Banking","row_ids":["02b14b2382923c4835faa17152c2d94b"],"score":96.15},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1700},"family":"GDPval-AA","row_ids":["022f63e78e1f4658e8af4b434e595fa3"],"score":62.96},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":160},"family":"Terminal-Bench","row_ids":["bb15f242cbb67b0a52ecb1efaeaf802e"],"score":53.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1100},"family":"AutomationBench","row_ids":["07fee1c32e6f2287a67ddf683ea02c91"],"score":57.89},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":700},"family":"AA-Briefcase","row_ids":["bee797f9fafa487f710b7abf6c3fddcf"],"score":63.64},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"EnterpriseOps-Gym","row_ids":["cea12af74b5609df87befba4292e8ab2"],"score":66.67}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","included":true,"model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2.1","peer_count":82,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","included":true,"model_id":"m-c947385433db3f03","original_version_label":"4.0","peer_count":76,"percentile":53.33,"percentile_exact":{"denominator":3,"numerator":160},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","included":true,"model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":78,"percentile":63.64,"percentile_exact":{"denominator":11,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA run","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471}],"score":66.77,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":21,"numerator":800},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":300},"family":"FrontierCode","row_ids":["328d104cb62d50eef3ae2e92cc4db522"],"score":42.86},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":100},"family":"DeepSWE","row_ids":["b993c11069821aca78f26acbce738dbc"],"score":33.33}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","included":true,"model_id":"m-c947385433db3f03","original_version_label":"2.0","peer_count":50,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","included":true,"model_id":"m-c947385433db3f03","original_version_label":"1.1","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2}],"score":38.1,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":9063,"numerator":536350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"IFBench","row_ids":["860832a18411b625ebef7842433113ca"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1300},"family":"GDP.pdf","row_ids":["4189c00054cffddacad0958de8d0f156"],"score":68.42},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["5ce6a4a59546a3ac7bd2230a498f7f48","ee80bfee31969adf86abcbeaae10d4e4"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current published protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742}],"score":59.18,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":600},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":600},"family":"MMMU-Pro","row_ids":["fa359632aae944ad7f900cc61aa55351"],"score":35.29}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821}],"score":35.29,"weight":10}],"eligible":true,"id":"m-c947385433db3f03","in_window":true,"name":"Qwen3.8 27B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-14","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":40,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","included":true,"model_id":"m-c947385433db3f03","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond","peer_count":64,"percentile":50.79,"percentile_exact":{"denominator":63,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","included":true,"model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2.1","peer_count":82,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","included":true,"model_id":"m-c947385433db3f03","original_version_label":"4.0","peer_count":76,"percentile":53.33,"percentile_exact":{"denominator":3,"numerator":160},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","included":true,"model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":78,"percentile":63.64,"percentile_exact":{"denominator":11,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA run","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","included":true,"model_id":"m-c947385433db3f03","original_version_label":"2.0","peer_count":50,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","included":true,"model_id":"m-c947385433db3f03","original_version_label":"1.1","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current published protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821}],"release_date":"2026-08-14","release_status":"published","release_verified":true,"result_updated_at":"2026-08-26","score":51.64,"score_exact":{"denominator":2620592406432,"numerator":135314940818345},"source_id":"alibaba/qwen3.8-27b"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":164300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2500},"family":"GPQA Diamond","row_ids":["4ac219e06b407a4da2953a84f21a360f"],"score":42.37},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2900},"family":"HLE","row_ids":["5706c727b7c4831ffe377f7ebcc17796"],"score":46.03}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":42.37,"percentile_exact":{"denominator":59,"numerator":2500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","peer_count":64,"percentile":46.03,"percentile_exact":{"denominator":63,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146}],"score":44.2,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":293031139},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":250},"family":"tau3-Banking","row_ids":["c4bd6d87d676b473fe7bbadf8b7238e4"],"score":19.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1100},"family":"GDPval-AA","row_ids":["0bf079be27f6b7285cbdd993ff5b5a93"],"score":13.58},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":178},"family":"Terminal-Bench","row_ids":["4473f9853fd2b2892365909edf9631fb"],"score":59.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":975},"family":"AutomationBench","row_ids":["20e66aebafb2226fab822f3a7524d54c"],"score":51.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3600},"family":"AA-Briefcase","row_ids":["110052f3248350cfda8c6580116d3dd9"],"score":46.75}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","peer_count":82,"percentile":13.58,"percentile_exact":{"denominator":81,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"4.0","peer_count":76,"percentile":59.33,"percentile_exact":{"denominator":3,"numerator":178},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":51.32,"percentile_exact":{"denominator":19,"numerator":975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","peer_count":78,"percentile":46.75,"percentile_exact":{"denominator":77,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66}],"score":38.04,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":1450},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1450},"family":"GDP.pdf","row_ids":["9ab89470db83472d1b5d2ed28ebc6783"],"score":76.32}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202}],"score":76.32,"weight":20}],"eligible":true,"id":"m-3fe24456aa661a33","in_window":true,"name":"GPT-5.5 Instant (June 2026)","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":41,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":42.37,"percentile_exact":{"denominator":59,"numerator":2500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","peer_count":64,"percentile":46.03,"percentile_exact":{"denominator":63,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","peer_count":82,"percentile":13.58,"percentile_exact":{"denominator":81,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"4.0","peer_count":76,"percentile":59.33,"percentile_exact":{"denominator":3,"numerator":178},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":51.32,"percentile_exact":{"denominator":19,"numerator":975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","peer_count":78,"percentile":46.75,"percentile_exact":{"denominator":77,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202}],"release_date":"2026-06-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":51.18,"score_exact":{"denominator":1272485214,"numerator":65122894901},"source_id":"published/aa:gpt-5-5-instant-06-26"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":800394,"numerator":45089825},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5350},"family":"GPQA Diamond","row_ids":["a4b58fb51111dbcc6b81f75ff3f7fc61"],"score":90.68},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["0898509511ccb710480bf98d3742bbb0"],"score":31.58},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1300},"family":"HLE","row_ids":["d543f8764a78abfabdc05d4f312b8b2f"],"score":61.9},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":700},"family":"CritPt","row_ids":["a2d7db240ac12326c1a9bfd85ff9b349"],"score":41.18}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","peer_count":64,"percentile":61.9,"percentile_exact":{"denominator":21,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","peer_count":52,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235}],"score":56.33,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":145236,"numerator":8347475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["76dde74584cddc5a2d8618786061726c"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1025},"family":"tau3-Banking","row_ids":["20bfc5477d34134d5c7c99539fd481a0"],"score":78.85},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"GDPval-AA","row_ids":["727c30106a7ec7d9e1c1205d514e6937"],"score":66.67},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":125},"family":"Terminal-Bench","row_ids":["35687651705594abdbf62890a0bd5a87"],"score":31.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1300},"family":"AutomationBench","row_ids":["63186f8028fad18deb2fa4c75b5a0888"],"score":68.42},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":400},"family":"AA-Briefcase","row_ids":["40fe84ac7f2d6389219a5c06f59da083"],"score":57.14},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["6c40f85770ab1fe03b45de2938f913b3"],"score":100.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":78.85,"percentile_exact":{"denominator":13,"numerator":1025},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","peer_count":82,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"4.0","peer_count":49,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":78,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731}],"score":57.48,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":441,"numerator":17000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2400},"family":"FrontierCode","row_ids":["2e337d0b2b65e72e819796e03cdc53e7"],"score":48.98},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"Vals Code Migration","row_ids":["7e542189106406c0f4ba769f7753c91e"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":50},"family":"DeepSWE","row_ids":["63840894bc67cdd6c5a9cd6e26e8b9d7"],"score":16.67}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"2.0","peer_count":50,"percentile":48.98,"percentile_exact":{"denominator":49,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":27,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7}],"score":38.55,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2485600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1000},"family":"GDP.pdf","row_ids":["c287cda1f48a055602b23132b6d39858"],"score":52.63},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":900},"family":"SimpleQA Verified","row_ids":["665cc51403349d6e59eed00558808309"],"score":52.94},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["3f8c786b8fc874f5d0f0ae1174917198","775332c591df3bb4219419f088bb5ff5"],"score":39.62}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429}],"score":48.4,"weight":20}],"eligible":true,"id":"m-f936349cbdb1def9","in_window":true,"name":"DeepSeek V4 Pro 0813","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-13","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":42,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","peer_count":64,"percentile":61.9,"percentile_exact":{"denominator":21,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","peer_count":52,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":78.85,"percentile_exact":{"denominator":13,"numerator":1025},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","peer_count":82,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"4.0","peer_count":49,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":78,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"2.0","peer_count":50,"percentile":48.98,"percentile_exact":{"denominator":49,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":27,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429}],"release_date":"2026-08-13","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":50.94,"score_exact":{"denominator":416912428296,"numerator":21235615018525},"source_id":"deepseek/deepseek-v4-pro-0813"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":27846,"numerator":1152575},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":50},"family":"FrontierMath","row_ids":["8462c2a29fbbd735e478ce804db0a935"],"score":5.56},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["9bdac139a55bab7bdafc52c1059a96f7"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":400},"family":"HLE","row_ids":["3c62c893b3f2b3d94df10089539fb20a"],"score":57.14},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2500},"family":"CritPt","row_ids":["e52491cdc38740f220e3951458e0d11e"],"score":49.02}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","peer_count":64,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","peer_count":52,"percentile":49.02,"percentile_exact":{"denominator":51,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27}],"score":41.39,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":13167,"numerator":878050},"families":[{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":200},"family":"Terminal-Bench","row_ids":["038de3c11daf5dbf02e5d1c9dd11ccd7"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1350},"family":"AutomationBench","row_ids":["7fa3942a02994b01f2c4e2365b0583bc"],"score":71.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4800},"family":"AA-Briefcase","row_ids":["bc12c3c00877323848c26ea730d935cc"],"score":62.34}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"4.0","peer_count":76,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":78,"percentile":62.34,"percentile_exact":{"denominator":77,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31}],"score":66.69,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":2200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2200},"family":"FrontierCode","row_ids":["c316e6a1765e379f023423782e08477b"],"score":44.9}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"2.0","peer_count":50,"percentile":44.9,"percentile_exact":{"denominator":49,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313}],"score":44.9,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2154950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1350},"family":"GDP.pdf","row_ids":["a6443ea346efcda5f00f1ce182e0d982"],"score":71.05},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":50},"family":"SimpleQA Verified","row_ids":["39e767a0ed9e71374bab635fd9e2ce92"],"score":2.94},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2750},"family":"AA-Omniscience","row_ids":["6656d06e85a16beb97db67e4960f17c9","8175922707070fb09367c5a6424e1c3a"],"score":51.89}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018}],"score":41.96,"weight":20}],"eligible":true,"id":"m-72b806dc67894cf3","in_window":true,"name":"Mistral Large 4 Preview","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":43,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","peer_count":64,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","peer_count":52,"percentile":49.02,"percentile_exact":{"denominator":51,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"4.0","peer_count":76,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":78,"percentile":62.34,"percentile_exact":{"denominator":77,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"2.0","peer_count":50,"percentile":44.9,"percentile_exact":{"denominator":49,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018}],"release_date":"2026-10-06","release_status":"published","release_verified":true,"result_updated_at":"2026-10-12","score":49.32,"score_exact":{"denominator":38864717892,"numerator":1916929064525},"source_id":"mistral/mistral-large-4"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":513,"numerator":7325},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":50},"family":"FrontierMath","row_ids":["3121b8aa930fa12d352b2910814de607"],"score":5.56},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":100},"family":"GPQA Diamond","row_ids":["6c2c7071889205aa8d51f6bb7f06fb84"],"score":33.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["90672f02fa6c8adab2a2801b7c2aedb3"],"score":3.95}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"score":14.28,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":72,"numerator":5375},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":325},"family":"Terminal-Bench","row_ids":["4ce54d74c2045140751a930d5f81e63e"],"score":40.63},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["e28fafc89bbfaddac23f5b76027e9f84"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":250},"family":"AA-Briefcase","row_ids":["59421edaa49a1a1c8543f29553d821c6"],"score":83.33}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","peer_count":49,"percentile":40.63,"percentile_exact":{"denominator":8,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855}],"score":74.65,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":1100},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":900},"family":"Vals Code Migration","row_ids":["3beb1575e4852b44427673ea326ae656"],"score":69.23},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["3bd3c3d2983a2fb1db83fc1872674955"],"score":100.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":27,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2}],"score":84.62,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":550},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":550},"family":"SimpleQA Verified","row_ids":["66b5a46f01ac8af0a7586871cb050ba1"],"score":32.35}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429}],"score":32.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"Chartography","row_ids":["e5a64d88a17b8e69890219a69e61c957"],"score":33.33}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9}],"score":33.33,"weight":10}],"eligible":true,"id":"m-99183a5b3b302b0d","in_window":true,"name":"DeepSeek V4.1 Flash","official_evidence":{"artifact_sha256":"df2d95745d8bc7620e1be8267ede37334adde98086da78ba1e49574f407a619a","basis":"official_release_page","date":"2026-09-10","sha256":"df2d95745d8bc7620e1be8267ede37334adde98086da78ba1e49574f407a619a","url":"https://www.deepseek.com/en/news/deepseek-v4-1-flash/"},"primary_profile_id":"native","profile_id":"native","rank":44,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","peer_count":49,"percentile":40.63,"percentile_exact":{"denominator":8,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":27,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9}],"release_date":"2026-09-10","release_status":"published","release_verified":true,"result_updated_at":"2026-10-13","score":48.96,"score_exact":{"denominator":3627936,"numerator":177623315},"source_id":"deepseek/deepseek-v4.1-flash"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":41.67,"evaluator_count":2,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["6e4fcbf667751f178fabfb4106350208"],"score":31.58},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":25},"family":"HLE","row_ids":["5a5d1e52735d8efc5b98d11b601491a8"],"score":25.0}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"Diamond","peer_count":17,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4}],"score":28.29,"weight":25},{"effective_weight_percent":41.67,"evaluator_count":2,"exact":{"denominator":16,"numerator":975},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":175},"family":"Terminal-Bench","row_ids":["73167c475c92b2f6b68a011a68a996ab"],"score":21.88},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"AA-Briefcase","row_ids":["0619cb6c1df6def76a850c294e50de63"],"score":100.0}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"v1.1","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162}],"score":60.94,"weight":25},{"effective_weight_percent":16.67,"evaluator_count":1,"exact":{"denominator":27,"numerator":1900},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1900},"family":"Humanity Sixth Sense","row_ids":["fcb2b9f9c2c7859d3369fef86db555f4"],"score":70.37}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","peer_count":28,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4}],"score":70.37,"weight":10}],"eligible":true,"id":"m-c0580b01221ed260","in_window":true,"name":"Muse Spark 1.3","official_evidence":{"artifact_sha256":"b221b103503442df9d920dbb3ff2002c3c9c86755300e3f9635018eaea8c2e56","basis":"official_release_page","date":"2026-09-02","sha256":"b221b103503442df9d920dbb3ff2002c3c9c86755300e3f9635018eaea8c2e56","url":"https://research.meta.ai/blog/introducing-muse-spark-1-3"},"primary_profile_id":"native","profile_id":"native","rank":45,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"Diamond","peer_count":17,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"v1.1","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","peer_count":28,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"result_updated_at":null,"score":48.91,"score_exact":{"denominator":98496,"numerator":4817075},"source_id":"meta/muse-spark-1.3"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":45.45,"evaluator_count":2,"exact":{"denominator":144,"numerator":6425},"families":[{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":275},"family":"Terminal-Bench","row_ids":["bd6d00a622ee8666fe845f21e2f0650f"],"score":11.46},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":700},"family":"AA-Briefcase","row_ids":["2a4003430f96eaca6cf7fbdac62ade55"],"score":77.78}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154}],"score":44.62,"weight":25},{"effective_weight_percent":36.36,"evaluator_count":1,"exact":{"denominator":17,"numerator":1200},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1200},"family":"SimpleQA Verified","row_ids":["aa61e58d244812473e1af5a8764b4e07"],"score":70.59}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844}],"score":70.59,"weight":20},{"effective_weight_percent":18.18,"evaluator_count":1,"exact":{"denominator":27,"numerator":400},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":400},"family":"Humanity Sixth Sense","row_ids":["b1617164f055d3d72ddaf46f34da89d8"],"score":14.81}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","peer_count":28,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5}],"score":14.81,"weight":10}],"eligible":true,"id":"m-2f22b07da3af863c","in_window":true,"name":"Muse Spark 1.2","official_evidence":{"artifact_sha256":"92a7e106c1f5836b322f76df67124f4f08c92c8dbe8cfa67e78698bd91fb844f","basis":"official_release_page","date":"2026-08-05","sha256":"92a7e106c1f5836b322f76df67124f4f08c92c8dbe8cfa67e78698bd91fb844f","url":"https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2"},"primary_profile_id":"native","profile_id":"native","rank":46,"records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","peer_count":28,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5}],"release_date":"2026-08-05","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":48.64,"score_exact":{"denominator":80784,"numerator":3929575},"source_id":"meta/muse-spark-1.2"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":469404,"numerator":20872015},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":200},"family":"FrontierMath","row_ids":["c2edb36eb77a36c2b7e3b5065aafdd4c"],"score":22.22},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4000},"family":"GPQA Diamond","row_ids":["bf4b9ff10398e116bbd0ed6eb4a0a88d"],"score":67.8},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["356d1ac12dc91c2d1cadb615b7e3023b"],"score":46.15},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":125},"family":"HLE","row_ids":["04518aae51e855cbdd25679b26d57f51"],"score":31.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["3eef34fe8a68f3f65cb23f07be80c5e4"],"score":54.9}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","included":true,"model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":67.8,"percentile_exact":{"denominator":59,"numerator":4000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","included":true,"model_id":"m-db208dec31031420","original_version_label":"0.1","peer_count":53,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":44.46,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":13945607},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5000},"family":"GDPval-AA","row_ids":["c140b9f2f3b8b047f022d22a6670f4c4"],"score":61.73},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":164},"family":"Terminal-Bench","row_ids":["6d1fe83c673cc3e53108a1f798f9857e"],"score":54.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1150},"family":"AutomationBench","row_ids":["cc297748bec17bb0b4aa16e861980849"],"score":60.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4500},"family":"AA-Briefcase","row_ids":["00f62480bdc3ab7dea00bbda96ff9331"],"score":58.44}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2.1","peer_count":82,"percentile":61.73,"percentile_exact":{"denominator":81,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","included":true,"model_id":"m-db208dec31031420","original_version_label":"4.0","peer_count":76,"percentile":54.67,"percentile_exact":{"denominator":3,"numerator":164},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","included":true,"model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":78,"percentile":58.44,"percentile_exact":{"denominator":77,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69}],"score":58.84,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":15950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1700},"family":"FrontierCode","row_ids":["ca9b3bebd5cf2b1f50a3f98180ff0076"],"score":34.69},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":200},"family":"Vals Code Migration","row_ids":["aeba1081b4abe0557cd65f41eca5f8b3"],"score":15.38}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","included":true,"model_id":"m-db208dec31031420","original_version_label":"2.0","peer_count":50,"percentile":34.69,"percentile_exact":{"denominator":49,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","included":true,"model_id":"m-db208dec31031420","original_version_label":"1.1","peer_count":27,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89}],"score":25.04,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":8496475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1375},"family":"GDP.pdf","row_ids":["7048e36b711cfdf67298d82bfa3d5f86"],"score":72.37},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1100},"family":"SimpleQA Verified","row_ids":["2fc7380a8d1e0cbfdc6837142943acdd"],"score":64.71},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"MultiChallenge","row_ids":["fc1a81b2311badda3a38cf9a6a28b5f4"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3900},"family":"AA-Omniscience","row_ids":["dbcce4096cefcfd79cea60ad99bf1516","0b6767461c06c4b95e61401653221d25"],"score":73.58}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","included":true,"model_id":"m-db208dec31031420","original_version_label":"current AA protocol","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","included":true,"model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","included":true,"model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412}],"score":62.04,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":24700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1000},"family":"Humanity Sixth Sense","row_ids":["fe8e2c3470907e5adda03a06d7b5f363"],"score":37.04},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1200},"family":"MMMU-Pro","row_ids":["71c04ca8ce2b90585afee7d563834b34"],"score":70.59}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","included":true,"model_id":"m-db208dec31031420","original_version_label":"current public protocol","peer_count":28,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","included":true,"model_id":"m-db208dec31031420","original_version_label":"current Standard 10","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295}],"score":53.81,"weight":10}],"eligible":true,"id":"m-db208dec31031420","in_window":true,"name":"Muse Spark Open","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":47,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","included":true,"model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":67.8,"percentile_exact":{"denominator":59,"numerator":4000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","included":true,"model_id":"m-db208dec31031420","original_version_label":"0.1","peer_count":53,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2.1","peer_count":82,"percentile":61.73,"percentile_exact":{"denominator":81,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","included":true,"model_id":"m-db208dec31031420","original_version_label":"4.0","peer_count":76,"percentile":54.67,"percentile_exact":{"denominator":3,"numerator":164},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","included":true,"model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":78,"percentile":58.44,"percentile_exact":{"denominator":77,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","included":true,"model_id":"m-db208dec31031420","original_version_label":"2.0","peer_count":50,"percentile":34.69,"percentile_exact":{"denominator":49,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","included":true,"model_id":"m-db208dec31031420","original_version_label":"1.1","peer_count":27,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","included":true,"model_id":"m-db208dec31031420","original_version_label":"current AA protocol","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","included":true,"model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","included":true,"model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","included":true,"model_id":"m-db208dec31031420","original_version_label":"current public protocol","peer_count":28,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","included":true,"model_id":"m-db208dec31031420","original_version_label":"current Standard 10","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"result_updated_at":"2026-10-15","score":48.62,"score_exact":{"denominator":9172073422512,"numerator":445978209583391},"source_id":"meta/muse-spark-open"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":13923,"numerator":703600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["e9c81f27a54c74df81f07409ddb33f17"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":300},"family":"HLE","row_ids":["433cbb1c2fcf649c36b004a451515cf8"],"score":42.86},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["737809004b185d439bc3bdb639d99a34"],"score":54.9}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","included":true,"model_id":"m-60cd64beda366768","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","included":true,"model_id":"m-60cd64beda366768","original_version_label":"Diamond","peer_count":64,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":50.54,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":1896048,"numerator":133400075},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6100},"family":"GDPval-AA","row_ids":["42544e43967725b8e648b9b8615120d6"],"score":75.31},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":575},"family":"Terminal-Bench","row_ids":["3fbf327bd9cbc7d6e3195a9d909b9231"],"score":47.92},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1550},"family":"AutomationBench","row_ids":["250157187c56354b4eed518f47426340"],"score":81.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5900},"family":"AA-Briefcase","row_ids":["e5699451024628001ea794ad1191793b"],"score":76.62}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2.1","peer_count":82,"percentile":75.31,"percentile_exact":{"denominator":81,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","included":true,"model_id":"m-60cd64beda366768","original_version_label":"4.0","peer_count":49,"percentile":47.92,"percentile_exact":{"denominator":12,"numerator":575},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","included":true,"model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":78,"percentile":76.62,"percentile_exact":{"denominator":77,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83}],"score":70.36,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":27425},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2900},"family":"FrontierCode","row_ids":["d9ee15576eab5d8aeca7fb1035d14154"],"score":59.18},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":350},"family":"Vals Code Migration","row_ids":["501724f84396d924c42fcaca484ef573"],"score":26.92}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","included":true,"model_id":"m-60cd64beda366768","original_version_label":"2.0","peer_count":50,"percentile":59.18,"percentile_exact":{"denominator":49,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","included":true,"model_id":"m-60cd64beda366768","original_version_label":"1.1","peer_count":27,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55}],"score":43.05,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":36975},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":750},"family":"GDP.pdf","row_ids":["0f9316648f89e3d521fa219b6361fd6c"],"score":39.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["084eb35df8b2a6c0ecbd814a3709a769","b4e27389bf022ba2e7fea370cfc23e71"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253}],"score":36.72,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":10000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":700},"family":"Humanity Sixth Sense","row_ids":["672a5ff05bade6e76dd2270c17366a35"],"score":25.93},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":300},"family":"MMMU-Pro","row_ids":["f2f6f962e776bf31934ea15f004e8009"],"score":17.65}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current public protocol","peer_count":28,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289}],"score":21.79,"weight":10}],"eligible":true,"id":"m-60cd64beda366768","in_window":true,"name":"MiMo-V2.6-Flash","official_evidence":{"artifact_sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","basis":"official_release_page","date":"2026-09-22","sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","url":"https://mimo.mi.com/docs/en-US/updates/model"},"primary_profile_id":"native","profile_id":"native","rank":48,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","included":true,"model_id":"m-60cd64beda366768","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","included":true,"model_id":"m-60cd64beda366768","original_version_label":"Diamond","peer_count":64,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2.1","peer_count":82,"percentile":75.31,"percentile_exact":{"denominator":81,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","included":true,"model_id":"m-60cd64beda366768","original_version_label":"4.0","peer_count":49,"percentile":47.92,"percentile_exact":{"denominator":12,"numerator":575},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","included":true,"model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":78,"percentile":76.62,"percentile_exact":{"denominator":77,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","included":true,"model_id":"m-60cd64beda366768","original_version_label":"2.0","peer_count":50,"percentile":59.18,"percentile_exact":{"denominator":49,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","included":true,"model_id":"m-60cd64beda366768","original_version_label":"1.1","peer_count":27,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current public protocol","peer_count":28,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":48.36,"score_exact":{"denominator":621835486272,"numerator":30069425541005},"source_id":"xiaomi/mimo-v2.6-flash"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1239,"numerator":75700},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4400},"family":"GPQA Diamond","row_ids":["7516363fe155a8bbdfc789cee4300773"],"score":74.58},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1000},"family":"HLE","row_ids":["7c461dbccbd2115e299061bd3cfb0f10"],"score":47.62}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":74.58,"percentile_exact":{"denominator":59,"numerator":4400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","peer_count":64,"percentile":47.62,"percentile_exact":{"denominator":21,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148}],"score":61.1,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":440154,"numerator":17585753},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":575},"family":"tau3-Banking","row_ids":["1b7ece54bb0355e69302fd5373585994"],"score":44.23},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1000},"family":"GDPval-AA","row_ids":["ae60755488af790d34ab9ed3aa726374"],"score":37.04},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["3126ca26ac0a6c7dab9a15033082575d"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":850},"family":"AutomationBench","row_ids":["2f9ec0e2d5074f31fb106a7f0db810e6"],"score":44.74},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":300},"family":"AA-Briefcase","row_ids":["8c6b16e6477c30b02c0cf413d3d98aff"],"score":27.27},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"EnterpriseOps-Gym","row_ids":["1d959564327d6d5496bf9174d774a4c8"],"score":44.44}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","peer_count":82,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","peer_count":78,"percentile":27.27,"percentile_exact":{"denominator":11,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996}],"score":39.95,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":2,"exact":{"denominator":3876,"numerator":157525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":225},"family":"IFBench","row_ids":["15ec6c548623aedd9b87ffa73167943d"],"score":56.25},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["60b6aea7e19a90a921731bc7d92e251c"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":300},"family":"SimpleQA Verified","row_ids":["29624d33497b02671ed8ec8a79e7f2e8"],"score":17.65}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365}],"score":40.64,"weight":20}],"eligible":true,"id":"m-5ea285ac7903cbc7","in_window":true,"name":"Kimi K2.7 Code","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":49,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":74.58,"percentile_exact":{"denominator":59,"numerator":4400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","peer_count":64,"percentile":47.62,"percentile_exact":{"denominator":21,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","peer_count":82,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","peer_count":78,"percentile":27.27,"percentile_exact":{"denominator":11,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365}],"release_date":"2026-06-12","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":47.7,"score_exact":{"denominator":43264497276,"numerator":2063781354715},"source_id":"moonshotai/kimi-k2.7-code"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":32214,"numerator":1462025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3600},"family":"GPQA Diamond","row_ids":["90978c171b499bb2efc92c2da0d369b4"],"score":61.02},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["5952fe944fd804ef75ab5466dc485304"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":250},"family":"HLE","row_ids":["7efb7f319954edefb37b62ce3d3ff826"],"score":35.71}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":61.02,"percentile_exact":{"denominator":59,"numerator":3600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116}],"score":45.38,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":4621617,"numerator":208057310},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"APEX-Agents","row_ids":["b5b0e9a07c5a8ced9fdf4c9ac74df7fc"],"score":50.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":775},"family":"tau3-Banking","row_ids":["84d4042d1c82bfee0ac491e53b4fa917"],"score":59.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4400},"family":"GDPval-AA","row_ids":["5a309e2224372571fd72b2e3de2fe6e1"],"score":54.32},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["cafc26cd33f16c1828c33bad26ea21a7"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":925},"family":"AutomationBench","row_ids":["2d900bebd8be5f37326880467236361d"],"score":48.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3400},"family":"AA-Briefcase","row_ids":["8d8e39fb86f2781e20fa7a7f68dc2ca4"],"score":44.16}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":59.62,"percentile_exact":{"denominator":13,"numerator":775},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","peer_count":82,"percentile":54.32,"percentile_exact":{"denominator":81,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":48.68,"percentile_exact":{"denominator":19,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","peer_count":78,"percentile":44.16,"percentile_exact":{"denominator":77,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16}],"score":45.02,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":38,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["ec9b7d6c1e90c4addeb4e33b7ca0403e"],"score":48.03}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"score":48.03,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":950},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":950},"family":"MMMU-Pro","row_ids":["d16dcb5ce825969190ea4f39de9f84b5"],"score":55.88}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867}],"score":55.88,"weight":10}],"eligible":true,"id":"m-d031501e2d0117d8","in_window":true,"name":"Apodex 1.1","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-30","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":50,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":61.02,"percentile_exact":{"denominator":59,"numerator":3600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":59.62,"percentile_exact":{"denominator":13,"numerator":775},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","peer_count":82,"percentile":54.32,"percentile_exact":{"denominator":81,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":48.68,"percentile_exact":{"denominator":19,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","peer_count":78,"percentile":44.16,"percentile_exact":{"denominator":77,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867}],"release_date":"2026-08-30","release_status":"published","release_verified":true,"result_updated_at":null,"score":47.24,"score_exact":{"denominator":148335419232,"numerator":7007785076525},"source_id":"published/aa:apodex-1-1"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":663,"numerator":45650},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["6de4c10b838de9bb43446d62c1dedc6c"],"score":65.38},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":200},"family":"HLE","row_ids":["ca6b622f3e2324e23758731c3eedb79c"],"score":66.67},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3800},"family":"CritPt","row_ids":["a69315d97d8db2867eda5787e5741424"],"score":74.51}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"0.1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"Diamond","peer_count":64,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","peer_count":52,"percentile":74.51,"percentile_exact":{"denominator":51,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425}],"score":68.85,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":32706828,"numerator":1477535075},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5200},"family":"GDPval-AA","row_ids":["f4a874da20b414e345395af6876bfede"],"score":64.2},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":175},"family":"Terminal-Bench","row_ids":["a8bde16affe89631785f57ec46935865"],"score":29.17},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1225},"family":"AutomationBench","row_ids":["c468f61930c57074fdae0429fd64cde2"],"score":64.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4700},"family":"AA-Briefcase","row_ids":["27c5dd58443826f50a40adf5e5fbdf05"],"score":61.04},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"OSWorld","row_ids":["b4bc853a9a856396a7313b82125218ee"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1200},"family":"AA-AnalystAgent","row_ids":["ebd2172910fbd66a3c4a6655aa5bdef4"],"score":52.17}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2.1","peer_count":82,"percentile":64.2,"percentile_exact":{"denominator":81,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","included":true,"model_id":"m-e5d076977e951607","original_version_label":"4.0","peer_count":49,"percentile":29.17,"percentile_exact":{"denominator":6,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","included":true,"model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":78,"percentile":61.04,"percentile_exact":{"denominator":77,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.1","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","peer_count":24,"percentile":52.17,"percentile_exact":{"denominator":23,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375}],"score":45.18,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":4,"exact":{"denominator":1274,"numerator":32475},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3300},"family":"FrontierCode","row_ids":["91532a2952c890cc776404a1ca00195c"],"score":67.35},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":450},"family":"Vals Code Migration","row_ids":["1045763a4cc14f13d71cbc23b34d8b21"],"score":34.62},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["12d12a117fed600435582eb0cdccab38"],"score":0.0},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":0},"family":"MirrorCode","row_ids":["9b168b30e93ee9c993180224bf1ba997"],"score":0.0}],"family_count":4,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.0","peer_count":50,"percentile":67.35,"percentile_exact":{"denominator":49,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":27,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666}],"score":25.49,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":1755175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3225},"family":"GDP.pdf","row_ids":["ff1f819163bf03455ec9cbe6d10872e3"],"score":84.87},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":450},"family":"SimpleQA Verified","row_ids":["d183987f58d811d89f8c42a38fe26df4"],"score":26.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["efcb1058be3817d672ab8961f07089a0","7ec06e89d9178a8c5c27508ed35805db"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906}],"score":51.26,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":51,"numerator":1000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Humanity Sixth Sense","row_ids":["c23775cd50aa5f64341fa34e8b1cf6d3"],"score":0.0},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1000},"family":"MMMU-Pro","row_ids":["7443158d8b6f105fcb7949451afd84ee"],"score":58.82},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["4a70f456ec596dc9670786d2fcc37f1a"],"score":0.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current public protocol","peer_count":28,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","included":true,"model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1}],"score":19.61,"weight":10}],"eligible":true,"id":"m-e5d076977e951607","in_window":true,"name":"GPT-6 Luna","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":51,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"0.1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"Diamond","peer_count":64,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","peer_count":52,"percentile":74.51,"percentile_exact":{"denominator":51,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2.1","peer_count":82,"percentile":64.2,"percentile_exact":{"denominator":81,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","included":true,"model_id":"m-e5d076977e951607","original_version_label":"4.0","peer_count":49,"percentile":29.17,"percentile_exact":{"denominator":6,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","included":true,"model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":78,"percentile":61.04,"percentile_exact":{"denominator":77,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.1","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","peer_count":24,"percentile":52.17,"percentile_exact":{"denominator":23,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.0","peer_count":50,"percentile":67.35,"percentile_exact":{"denominator":49,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":27,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current public protocol","peer_count":28,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","included":true,"model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":"2026-10-26","score":45.82,"score_exact":{"denominator":10726662138192,"numerator":491483807341325},"source_id":"openai/gpt-6-luna"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":2457,"numerator":80425},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":575},"family":"Terminal-Bench-Science","row_ids":["63428ab4d6f5aa0cb6c23626f378d1a8"],"score":44.23},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1300},"family":"HLE","row_ids":["5c428eec7c6c7eddbfd00d536e4c24e0"],"score":20.63},{"benchmark_version":"2 Hard","exact":{"denominator":3,"numerator":100},"family":"CritPt","row_ids":["2147a64a171deda69e1dcacdea28ad28"],"score":33.33}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"0.1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","peer_count":64,"percentile":20.63,"percentile_exact":{"denominator":63,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","peer_count":52,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155}],"score":32.73,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":4389,"numerator":275636},"families":[{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":72},"family":"Terminal-Bench","row_ids":["37a1dedbc40bda682abb927c484863ce"],"score":72.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1200},"family":"AutomationBench","row_ids":["87eb4175855e40447de186dd2f3411a6"],"score":63.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4100},"family":"AA-Briefcase","row_ids":["24d71c9e7d3f6a3d2ddf4b46e341dafa"],"score":53.25}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"4.0","peer_count":76,"percentile":72.0,"percentile_exact":{"denominator":1,"numerator":72},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":78,"percentile":53.25,"percentile_exact":{"denominator":77,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84}],"score":62.8,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1800},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1800},"family":"FrontierCode","row_ids":["9302ab59d95f9c2a82c7d7c38865f23b"],"score":36.73}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"2.0","peer_count":50,"percentile":36.73,"percentile_exact":{"denominator":49,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202}],"score":36.73,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":195525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["632aef0c9b2c1ee60a8fee169a7ead99"],"score":48.03},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2600},"family":"AA-Omniscience","row_ids":["464c122b5f2d778a04114dc266e1a6b8","d5ec18e90fffd664c6089d00e270f823"],"score":49.06}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698}],"score":48.54,"weight":20}],"eligible":true,"id":"m-8555d29b166d94e4","in_window":true,"name":"Nanbeige4.3-33B","official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-10-25","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":52,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"0.1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","peer_count":64,"percentile":20.63,"percentile_exact":{"denominator":63,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","peer_count":52,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"4.0","peer_count":76,"percentile":72.0,"percentile_exact":{"denominator":1,"numerator":72},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":78,"percentile":53.25,"percentile_exact":{"denominator":77,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"2.0","peer_count":50,"percentile":36.73,"percentile_exact":{"denominator":49,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698}],"release_date":"2026-10-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":45.49,"score_exact":{"denominator":1714619907,"numerator":77994001630},"source_id":"published/aa:nanbeige4-3-33b"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":730184,"numerator":43081725},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5500},"family":"GPQA Diamond","row_ids":["67157f1183905451cf4dded1391fab8e"],"score":93.22},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["1a6a33cd4cc454e60378e0a0694263f3"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1100},"family":"HLE","row_ids":["4912807a823c5e919553ca48ade2b67a"],"score":52.38},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2600},"family":"CritPt","row_ids":["3d5749511482d0fbdaddfa896055ac6a"],"score":50.98}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":93.22,"percentile_exact":{"denominator":59,"numerator":5500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond","peer_count":64,"percentile":52.38,"percentile_exact":{"denominator":21,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","peer_count":52,"percentile":50.98,"percentile_exact":{"denominator":51,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28}],"score":59.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":992107116,"numerator":27805451125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":375},"family":"tau3-Banking","row_ids":["0494d5d3e701ac0729bd55524064b2fa"],"score":28.85},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4600},"family":"GDPval-AA","row_ids":["7ee504c210f7e12209fb9a5011615203"],"score":56.79},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":25},"family":"Terminal-Bench","row_ids":["8f361f3f1d062dfe3e449864fa84aec9"],"score":6.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":775},"family":"AutomationBench","row_ids":["c964aa700a761259aa3ab0adef555cba"],"score":40.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3700},"family":"AA-Briefcase","row_ids":["703518382f29bb0a80ddba827b81e22d"],"score":48.05},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["df3f56d8282f0ce2c7d297df530d227a"],"score":11.11},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":100},"family":"AA-AnalystAgent","row_ids":["8cef68049263078fc3646fc52bd5e75a"],"score":4.35}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","included":true,"model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":28.85,"percentile_exact":{"denominator":13,"numerator":375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2.1","peer_count":82,"percentile":56.79,"percentile_exact":{"denominator":81,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","included":true,"model_id":"m-da421acad97d607b","original_version_label":"4.0","peer_count":49,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","included":true,"model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":78,"percentile":48.05,"percentile_exact":{"denominator":77,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA run","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","peer_count":24,"percentile":4.35,"percentile_exact":{"denominator":23,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"score":28.03,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":21950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3000},"family":"FrontierCode","row_ids":["3f103c270013e8e3a189c86aaaac75c3"],"score":61.22},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":100},"family":"Vals Code Migration","row_ids":["fbec8a823687d43757fc0efa8ad26423"],"score":7.69}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"2.0","peer_count":50,"percentile":61.22,"percentile_exact":{"denominator":49,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","included":true,"model_id":"m-da421acad97d607b","original_version_label":"1.1","peer_count":27,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9}],"score":34.46,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":3021,"numerator":181675},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["89b2619aee2f39a13fb5f1677cbe7c28"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":775},"family":"GDP.pdf","row_ids":["7e0fa158109c3646bd13a80e33ec79ee"],"score":40.79},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["253af28f0e697d86e6b8e62f78eae062","e14051d6dbdd6fd61ec4748398bf8936"],"score":39.62}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043}],"score":60.14,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":750},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":750},"family":"MMMU-Pro","row_ids":["9161d7aea787f48463e1dc1ca6952542"],"score":44.12}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977}],"score":44.12,"weight":10}],"eligible":true,"id":"m-da421acad97d607b","in_window":true,"name":"MiniMax-M3","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":53,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":93.22,"percentile_exact":{"denominator":59,"numerator":5500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond","peer_count":64,"percentile":52.38,"percentile_exact":{"denominator":21,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","peer_count":52,"percentile":50.98,"percentile_exact":{"denominator":51,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","included":true,"model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":28.85,"percentile_exact":{"denominator":13,"numerator":375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2.1","peer_count":82,"percentile":56.79,"percentile_exact":{"denominator":81,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","included":true,"model_id":"m-da421acad97d607b","original_version_label":"4.0","peer_count":49,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","included":true,"model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":78,"percentile":48.05,"percentile_exact":{"denominator":77,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA run","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","peer_count":24,"percentile":4.35,"percentile_exact":{"denominator":23,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"2.0","peer_count":50,"percentile":61.22,"percentile_exact":{"denominator":49,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","included":true,"model_id":"m-da421acad97d607b","original_version_label":"1.1","peer_count":27,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977}],"release_date":"2026-06-01","release_status":"published","release_verified":true,"result_updated_at":null,"score":45.09,"score_exact":{"denominator":32455029033504,"numerator":1463328516402595},"source_id":"minimax/MiniMax-M3"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":2800},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2800},"family":"GPQA Diamond","row_ids":["8763ad1982a06ca4a70e322ffe1ddbd4"],"score":47.46}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":47.46,"percentile_exact":{"denominator":59,"numerator":2800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434}],"score":47.46,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":243243,"numerator":13098550},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":950},"family":"tau3-Banking","row_ids":["479061fd58fe09f756e592ab99fa1642"],"score":73.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3800},"family":"GDPval-AA","row_ids":["d1a5193baa7d56820cb5874aa60b7c4a"],"score":46.91},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3200},"family":"AA-Briefcase","row_ids":["56f3576f6b7e7bada3d4ea81a3e974d2"],"score":41.56}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","peer_count":82,"percentile":46.91,"percentile_exact":{"denominator":81,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","peer_count":78,"percentile":41.56,"percentile_exact":{"denominator":77,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75}],"score":53.85,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":500},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":500},"family":"GDP.pdf","row_ids":["b0db480f387627aef52b7f9b3d6804ac"],"score":26.32}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052}],"score":26.32,"weight":20}],"eligible":true,"id":"m-9e1e0f605d47cf02","in_window":true,"name":"Motif 3","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":54,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":47.46,"percentile_exact":{"denominator":59,"numerator":2800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","peer_count":82,"percentile":46.91,"percentile_exact":{"denominator":81,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","peer_count":78,"percentile":41.56,"percentile_exact":{"denominator":77,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.7,"score_exact":{"denominator":1908727821,"numerator":83411342375},"source_id":"motif-technologies/motif-3"},{"benchmark_count":3,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":50.0,"evaluator_count":1,"exact":{"denominator":59,"numerator":2300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2300},"family":"GPQA Diamond","row_ids":["29813bdc52dc98722342b740c2f2d3f2"],"score":38.98}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":38.98,"percentile_exact":{"denominator":59,"numerator":2300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505}],"score":38.98,"weight":25},{"effective_weight_percent":50.0,"evaluator_count":1,"exact":{"denominator":234,"numerator":11275},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":675},"family":"tau3-Banking","row_ids":["76481bf229eeeddb8b44a15648197449"],"score":51.92},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":400},"family":"GDPval-AA","row_ids":["5d726bb21290d73d92a00bf9993c5a4a"],"score":44.44}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":51.92,"percentile_exact":{"denominator":13,"numerator":675},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","peer_count":82,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37}],"score":48.18,"weight":25}],"eligible":true,"id":"m-af0edba2f76c80d6","in_window":true,"name":"G9v3-39A5B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-20","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":55,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":38.98,"percentile_exact":{"denominator":59,"numerator":2300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":51.92,"percentile_exact":{"denominator":13,"numerator":675},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","peer_count":82,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37}],"release_date":"2026-08-20","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.58,"score_exact":{"denominator":27612,"numerator":1203425},"source_id":"published/aa:g9v3-39a5b"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":38.46,"evaluator_count":2,"exact":{"denominator":24,"numerator":1025},"families":[{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":75},"family":"Terminal-Bench","row_ids":["a49b610cdf02973ee5519a3d58738257"],"score":18.75},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":200},"family":"AA-Briefcase","row_ids":["99d362c58056009c0378c89776bd8bfa"],"score":66.67}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","peer_count":49,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071}],"score":42.71,"weight":25},{"effective_weight_percent":30.77,"evaluator_count":2,"exact":{"denominator":4,"numerator":175},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["f4858ed8f9e084fde242d77a18dddeed"],"score":62.5},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":25},"family":"MirrorCode","row_ids":["09300d4eee5c7d673f52949c9b3dcb98"],"score":25.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224}],"score":43.75,"weight":20},{"effective_weight_percent":30.77,"evaluator_count":1,"exact":{"denominator":17,"numerator":750},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":750},"family":"SimpleQA Verified","row_ids":["5fb5dec681fa13f63b1b4d2e3c76490c"],"score":44.12}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483}],"score":44.12,"weight":20}],"eligible":true,"id":"m-b0c41bb29e51489d","in_window":true,"name":"Grok 4.5","official_evidence":{"artifact_sha256":"605d18dafa601ac99d2ffe30c675010636f4c489483187f70c548f59080af7bf","basis":"official_release_page","date":"2026-07-16","sha256":"605d18dafa601ac99d2ffe30c675010636f4c489483187f70c548f59080af7bf","url":"https://x.ai/news/grok-4-5"},"primary_profile_id":"native","profile_id":"native","rank":56,"records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","peer_count":49,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483}],"release_date":"2026-07-16","release_status":"published","release_verified":true,"result_updated_at":"2026-10-21","score":43.46,"score_exact":{"denominator":5304,"numerator":230525},"source_id":"xai/grok-4.5"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":3100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3100},"family":"GPQA Diamond","row_ids":["012394f027c8418fe233e5f26b558306"],"score":52.54}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":52.54,"percentile_exact":{"denominator":59,"numerator":3100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545}],"score":52.54,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":118503,"numerator":6339076},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4700},"family":"GDPval-AA","row_ids":["87c313476c9a420ad27b41db7d0250bf"],"score":58.02},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":154},"family":"Terminal-Bench","row_ids":["8ad3da1ee13871ee0748612b7b0f2202"],"score":51.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1050},"family":"AutomationBench","row_ids":["f3bd4620346190964b5a6a49765359b6"],"score":55.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3800},"family":"AA-Briefcase","row_ids":["6f61748806659fcf7e221477d42b6563"],"score":49.35}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","peer_count":82,"percentile":58.02,"percentile_exact":{"denominator":81,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","peer_count":78,"percentile":49.35,"percentile_exact":{"denominator":77,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44}],"score":53.49,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":725},"family":"GDP.pdf","row_ids":["c038eca51adef7673b4117a8c612883d"],"score":38.16}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086}],"score":38.16,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":100},"family":"MMMU-Pro","row_ids":["90c2787f5b2a0845d431bc6af09c8c25"],"score":5.88}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746}],"score":5.88,"weight":10}],"eligible":true,"id":"m-2d46f5cc91cc119d","in_window":true,"name":"JT-4.1 Flash 236B A21B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-09","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":57,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":52.54,"percentile_exact":{"denominator":59,"numerator":3100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","peer_count":82,"percentile":58.02,"percentile_exact":{"denominator":81,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","peer_count":78,"percentile":49.35,"percentile_exact":{"denominator":77,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.41,"score_exact":{"denominator":475434036,"numerator":20638975985},"source_id":"published/aa:jt-4-1-flash-236b-a21b"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":115841125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3900},"family":"GPQA Diamond","row_ids":["b588de5f182f53797aeaba57de63e608"],"score":66.1},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["2f6e14dbb9becc8bf0feb78f2973083f"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1900},"family":"HLE","row_ids":["03d73659cb7b3abbbd4921c8546a8418"],"score":30.16},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1400},"family":"CritPt","row_ids":["5f0a8e1454688084cb0e096162e1eff2"],"score":27.45}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":66.1,"percentile_exact":{"denominator":59,"numerator":3900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond","peer_count":64,"percentile":30.16,"percentile_exact":{"denominator":63,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","peer_count":52,"percentile":27.45,"percentile_exact":{"denominator":51,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12}],"score":35.25,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":446904872},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":900},"family":"tau3-Banking","row_ids":["348ef206dd0fa625708f257a200455f5"],"score":69.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4900},"family":"GDPval-AA","row_ids":["3637cd472b3e658153b3601c1c953f3e"],"score":60.49},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":48},"family":"Terminal-Bench","row_ids":["612528520707bc7ebd5f35810f7efeb6"],"score":48.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1000},"family":"AutomationBench","row_ids":["d30d4db6bfda99c8317f5eddf5d0e239"],"score":52.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4600},"family":"AA-Briefcase","row_ids":["4c2ed1cac020a530bd05f7d661a8c39c"],"score":59.74}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2.1","peer_count":82,"percentile":60.49,"percentile_exact":{"denominator":81,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"4.0","peer_count":76,"percentile":48.0,"percentile_exact":{"denominator":1,"numerator":48},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":78,"percentile":59.74,"percentile_exact":{"denominator":77,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48}],"score":58.02,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1900},"family":"FrontierCode","row_ids":["b9e8a0463a865f6221caaaf1d3aecc9d"],"score":38.78}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"2.0","peer_count":50,"percentile":38.78,"percentile_exact":{"denominator":49,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465}],"score":38.78,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1007,"numerator":40025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":650},"family":"GDP.pdf","row_ids":["65f6a64cea971ea25fa466fbee3340a6"],"score":34.21},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2400},"family":"AA-Omniscience","row_ids":["d414f5fe5dd8b0e7a11df833233b1edf","7d19cde6db2f8d3d5e0cb451c988a92a"],"score":45.28}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506}],"score":39.75,"weight":20}],"eligible":true,"id":"m-d2febb93118535a2","in_window":true,"name":"K2 Horizon 375B A23B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":58,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":66.1,"percentile_exact":{"denominator":59,"numerator":3900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond","peer_count":64,"percentile":30.16,"percentile_exact":{"denominator":63,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","peer_count":52,"percentile":27.45,"percentile_exact":{"denominator":51,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2.1","peer_count":82,"percentile":60.49,"percentile_exact":{"denominator":81,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"4.0","peer_count":76,"percentile":48.0,"percentile_exact":{"denominator":1,"numerator":48},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":78,"percentile":59.74,"percentile_exact":{"denominator":77,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"2.0","peer_count":50,"percentile":38.78,"percentile_exact":{"denominator":49,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.36,"score_exact":{"denominator":41274330401304,"numerator":1789607818257619},"source_id":"published/aa:k2-horizon-375b-a23b"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":342,"numerator":12275},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":250},"family":"GPQA Diamond","row_ids":["dd3025c1130213f154e327d038efe0ed"],"score":83.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["f2b3352d05d0abb45473dcad237d2a14"],"score":11.84},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":25},"family":"HLE","row_ids":["0e05cbe602caeb64b25d2230f75c2113"],"score":12.5}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","peer_count":17,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2}],"score":35.89,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":72,"numerator":3887},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":80},"family":"APEX-Agents","row_ids":["e4b1483bcb889a5ef567c4c02fc571dd"],"score":80.0},{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":275},"family":"Terminal-Bench","row_ids":["fff5c31ca1adab72967ff7c824d4653e"],"score":34.38},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["cf571d5463813d8daec30e8a8f31740d"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":100},"family":"AA-Briefcase","row_ids":["04f715495ca6c55181e1b42a64312d78"],"score":11.11},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":400},"family":"MCP Atlas","row_ids":["b1436c94b33fa58416664a5d31b199a9"],"score":44.44}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3}],"score":53.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":450},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":250},"family":"Vals Code Migration","row_ids":["549f9102d89b3e9a3355e2ef41ef0184"],"score":19.23},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"DeepSWE","row_ids":["c87699e8afdc2f67ab86923665dd9ed5"],"score":50.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","peer_count":27,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5}],"score":34.62,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":50},"family":"SimpleQA Verified","row_ids":["63b449d94d8f27d0e405ea5ec8384866"],"score":50.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506}],"score":50.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":81,"numerator":2600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":800},"family":"Humanity Sixth Sense","row_ids":["d100a26c5dab08a99d576ce25ec3994e"],"score":29.63},{"benchmark_version":"Standard 10","exact":{"denominator":3,"numerator":200},"family":"MMMU-Pro","row_ids":["9ea024ab5ca2b01fdc71c3f50f184c9c"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["24019ec9f1eb6cf5c33ae9562d1f1bac"],"score":0.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","peer_count":28,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1}],"score":32.1,"weight":10}],"eligible":true,"id":"m-9e460c49820cbe4e","in_window":true,"name":"Kimi K3","official_evidence":{"artifact_sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","basis":"official_release_page","date":"2026-07-16","sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","url":"https://www.kimi.com/en/blog/"},"primary_profile_id":"native","profile_id":"native","rank":59,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","peer_count":17,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","peer_count":27,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","peer_count":28,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1}],"release_date":"2026-07-16","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":42.6,"score_exact":{"denominator":640224,"numerator":27275101},"source_id":"moonshotai/kimi-k3"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":531,"numerator":21650},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3500},"family":"GPQA Diamond","row_ids":["c2e927593d58bea9e466eb1677b5bc62"],"score":59.32},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":200},"family":"HLE","row_ids":["4fb23f5b07fa697e50fa64d17b095add"],"score":22.22}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":59.32,"percentile_exact":{"denominator":59,"numerator":3500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond","peer_count":64,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076}],"score":40.77,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":64571609},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":925},"family":"tau3-Banking","row_ids":["a5800dd77226525d5b0005d3b9d5c59c"],"score":71.15},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4100},"family":"GDPval-AA","row_ids":["2d97f2a57324a678b4a67b9a04a099de"],"score":50.62},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["b740ee2b3c6f41e732abe95d92f63313"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":650},"family":"AutomationBench","row_ids":["1aac376daa38513de1520f91177fbd00"],"score":34.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3100},"family":"AA-Briefcase","row_ids":["3a94cd016eb68403ef51f517cb4b5722"],"score":40.26}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":71.15,"percentile_exact":{"denominator":13,"numerator":925},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v2.1","peer_count":82,"percentile":50.62,"percentile_exact":{"denominator":81,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v1.1","peer_count":78,"percentile":40.26,"percentile_exact":{"denominator":77,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95}],"score":41.91,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":700},"family":"GDP.pdf","row_ids":["8a1843eddbf9ae3836989a362277ba86"],"score":36.84}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084}],"score":36.84,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":800},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":800},"family":"MMMU-Pro","row_ids":["487c57fe9e6c668745daa6efed20c475"],"score":47.06}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543}],"score":47.06,"weight":10}],"eligible":true,"id":"m-724c2a784624afa2","in_window":true,"name":"Ling-3.0-flash-VL","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":60,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":59.32,"percentile_exact":{"denominator":59,"numerator":3500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond","peer_count":64,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":71.15,"percentile_exact":{"denominator":13,"numerator":925},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v2.1","peer_count":82,"percentile":50.62,"percentile_exact":{"denominator":81,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v1.1","peer_count":78,"percentile":40.26,"percentile_exact":{"denominator":77,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543}],"release_date":"2026-09-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":40.93,"score_exact":{"denominator":24722569872,"numerator":1011958813885},"source_id":"inclusionai/ling-3.0-flash-vl"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7,"numerator":400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":400},"family":"GPQA Diamond","row_ids":["11f1be50c8d04272ce6fde6b77e7d32d"],"score":57.14}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"reference_revision":"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3}],"score":57.14,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":3,"exact":{"denominator":21,"numerator":550},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"APEX-Agents","row_ids":["7155f49a20e936c90be455db70c3df42"],"score":33.33},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":50},"family":"Terminal-Bench","row_ids":["54a707812fad1f7aeac89c3fb45c1b29"],"score":16.67},{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":200},"family":"AutomationBench","row_ids":["f68fab6bf158dd2600b8426c509f20a3"],"score":28.57}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"4.0","peer_count":49,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"reference_revision":"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1}],"score":26.19,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":2,"exact":{"denominator":91,"numerator":3350},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":400},"family":"Vals Code Migration","row_ids":["8b32286773fba058c9bb87cdd6ac0c71"],"score":30.77},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":300},"family":"DeepSWE","row_ids":["49c0ea386ce0f4eee60fb157edec7d0d"],"score":42.86}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"1.1","peer_count":27,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3}],"score":36.81,"weight":20}],"eligible":true,"id":"m-383f40728e3e5259","in_window":true,"name":"Hy4 preview","official_evidence":{"artifact_sha256":"676214cf8f2ac19b63d1293254a5e1920f5c5b758546dd77d7bbb0d132589ea3","basis":"official_release_page","date":"2026-08-28","sha256":"676214cf8f2ac19b63d1293254a5e1920f5c5b758546dd77d7bbb0d132589ea3","url":"https://www.tencent.com/zh-hk/tencent-releases-and-open-sources-tencent-hy4-preview/"},"primary_profile_id":"native","profile_id":"native","rank":61,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"reference_revision":"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"4.0","peer_count":49,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"reference_revision":"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"1.1","peer_count":27,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3}],"release_date":"2026-08-28","release_status":"published","release_verified":true,"result_updated_at":null,"score":40.28,"score_exact":{"denominator":1911,"numerator":76975},"source_id":"tencent/hy4-preview"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":531,"numerator":32500},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4600},"family":"GPQA Diamond","row_ids":["b8284820623630cb7a5150e367b7bed4"],"score":77.97},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":400},"family":"HLE","row_ids":["00361f2a88ddce87fbff29f4d32f2681"],"score":44.44}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":77.97,"percentile_exact":{"denominator":59,"numerator":4600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","peer_count":64,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139}],"score":61.21,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":13864851,"numerator":337939913},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["274b775cdefb96d48e1e4a5a4e4f52de"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["9202014a7bc9d24df89fe2534ffc35bc"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1900},"family":"GDPval-AA","row_ids":["c77b3a8d70f27e7507bc297339631f46"],"score":23.46},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["a84f629da93a24a910222138960c0c05"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":675},"family":"AutomationBench","row_ids":["1915dcb5be7d210d57854519bafaa4ef"],"score":35.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2600},"family":"AA-Briefcase","row_ids":["397f726491d251a65df1f043233be5de"],"score":33.77},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["e323a2908ddfad0f233fccef67ea0b29"],"score":50.0},{"benchmark_version":"4","exact":{"denominator":1,"numerator":0},"family":"BFCL","row_ids":["355d178390cc7668a17845fc6539c179"],"score":0.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["a05fb15039bcbd3145b6e77f10451f52"],"score":0.0}],"family_count":9,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","peer_count":82,"percentile":23.46,"percentile_exact":{"denominator":81,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","peer_count":78,"percentile":33.77,"percentile_exact":{"denominator":77,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2}],"score":24.37,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["9bf5f2dc87f09a1babb07763838a2e08"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":152,"numerator":9275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":175},"family":"IFBench","row_ids":["c4485d942b9b21c7a152b1888b11b7c4"],"score":87.5},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1075},"family":"GDP.pdf","row_ids":["850da6d9475999744f930f21d7861dd9"],"score":56.58},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"IFEval","row_ids":["5ec142fbe451bd07f915ecd120f2faec"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"MMLU-Pro","row_ids":["7be3425578af009ce953f32bc09ba9aa"],"score":0.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5}],"score":61.02,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1100},"family":"MMMU-Pro","row_ids":["b6422caa44d095e13cde40b5c1a62bc3"],"score":64.71}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647}],"score":64.71,"weight":10}],"eligible":true,"id":"m-7190ddc26a5a37a0","in_window":true,"name":"Qwen3.7 Plus","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":62,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":77.97,"percentile_exact":{"denominator":59,"numerator":4600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","peer_count":64,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","peer_count":82,"percentile":23.46,"percentile_exact":{"denominator":81,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","peer_count":78,"percentile":33.77,"percentile_exact":{"denominator":77,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647}],"release_date":"2026-06-01","release_status":"published","release_verified":true,"result_updated_at":"2026-08-26","score":40.07,"score_exact":{"denominator":111251564424,"numerator":4457774361283},"source_id":"alibaba/qwen3.7-plus"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":365092,"numerator":10439225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3000},"family":"GPQA Diamond","row_ids":["f63fe95853d236ed53d2b79d02b859cb"],"score":50.85},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["2c2a68b20b30379e3063400c424fa6d0"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":200},"family":"HLE","row_ids":["cda0d514d2e746d7e415df4602a30e61"],"score":28.57},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":300},"family":"CritPt","row_ids":["1b14f3e78b8a7ccb2a6b6d0eda26f9e2"],"score":17.65}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":50.85,"percentile_exact":{"denominator":59,"numerator":3000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","peer_count":64,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","peer_count":52,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035}],"score":28.59,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1027026,"numerator":39546907},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["62cc0b9af0e4241c44026a470cdd621c"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":100},"family":"GDPval-AA","row_ids":["9d93ac6bf71dd0301098e7d1b1db2fb3"],"score":33.33},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["54cd2ea0dc02b5047f97de0db07066fb"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":875},"family":"AutomationBench","row_ids":["bdc1d1c2838fa6785d600f455b24f788"],"score":46.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1500},"family":"AA-Briefcase","row_ids":["0ec16dc40db0414f2891438008ea0a1a"],"score":19.48},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":500},"family":"EnterpriseOps-Gym","row_ids":["5ffca0a973b63389e8c0ae797941349a"],"score":55.56}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","peer_count":82,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":46.05,"percentile_exact":{"denominator":19,"numerator":875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":78,"percentile":19.48,"percentile_exact":{"denominator":77,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188}],"score":38.51,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":1500},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1500},"family":"FrontierCode","row_ids":["26e16bd219d1e573463da6c55739a16d"],"score":30.61}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","peer_count":50,"percentile":30.61,"percentile_exact":{"denominator":49,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323}],"score":30.61,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":119275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1175},"family":"GDP.pdf","row_ids":["b1d8307905906d9eadee1409ffb734e6"],"score":61.84},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3000},"family":"AA-Omniscience","row_ids":["5ac3b36a409b29ee98389b739f2f25b7","4347afed89affe307401c55cb1f477b8"],"score":56.6}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043}],"score":59.22,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["ef305cf290c211e130604e205ee97f46"],"score":50.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624}],"score":50.0,"weight":10}],"eligible":true,"id":"m-e7f69d0dac9fe632","in_window":true,"name":"Gemini 3.5 Flash-Lite","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":63,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":50.85,"percentile_exact":{"denominator":59,"numerator":3000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","peer_count":64,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","peer_count":52,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","peer_count":82,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":46.05,"percentile_exact":{"denominator":19,"numerator":875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":78,"percentile":19.48,"percentile_exact":{"denominator":77,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","peer_count":50,"percentile":30.61,"percentile_exact":{"denominator":49,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624}],"release_date":"2026-07-21","release_status":"published","release_verified":true,"result_updated_at":null,"score":39.74,"score_exact":{"denominator":3057357807504,"numerator":121505356698007},"source_id":"google/gemini-3.5-flash-lite"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["1632686674422c3a2065b7adb4e89430"],"score":31.58}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":31.58,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":200},"families":[{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":200},"family":"AA-Briefcase","row_ids":["ec2017cdf9726d6330e3b72963cbf545"],"score":22.22}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808}],"score":22.22,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["2125ffdac9125e27a313037f91e5edb9"],"score":25.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0}],"score":25.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1325},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1325},"family":"SimpleQA Verified","row_ids":["8f6725165e10f2ce17308791e15f9d1c"],"score":77.94}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662}],"score":77.94,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":500},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":500},"family":"Humanity Sixth Sense","row_ids":["6d9db035c8e86692a9eb736f0a2c02ca"],"score":55.56}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","peer_count":28,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9}],"score":55.56,"weight":10}],"eligible":true,"id":"m-77b5dbc317075e9d","in_window":true,"name":"Gemini 3.6 Flash","official_evidence":{"artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","basis":"official_release_page","date":"2026-07-21","sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/"},"primary_profile_id":"native","profile_id":"native","rank":64,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","peer_count":28,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9}],"release_date":"2026-07-21","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":39.59,"score_exact":{"denominator":2907,"numerator":115100},"source_id":"google/gemini-3.6-flash"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":850},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["72ae1d1de2eb58dbc8e135c9056f08ba"],"score":44.74}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"score":44.74,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":96,"numerator":2275},"families":[{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":275},"family":"Terminal-Bench","row_ids":["349a8da908aa998fbc2c641cf58cfc1b"],"score":11.46},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["d209f552e4c2026bb776833fb9d113ca"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":50},"family":"AA-Briefcase","row_ids":["a4acfc86f781d098a642bc8bcfd57f42"],"score":16.67},{"benchmark_version":"1000 tasks","exact":{"denominator":3,"numerator":200},"family":"MCP Atlas","row_ids":["44434b3879169ad5c54a4316dc8731a9"],"score":66.67}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6}],"score":23.7,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["7b14fb9c23111ece0c4ecf6bf486054c"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1325},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1325},"family":"SimpleQA Verified","row_ids":["f90868b4246f3217c1fa61fa9b95b789"],"score":77.94}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662}],"score":77.94,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":27,"numerator":1600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1600},"family":"Humanity Sixth Sense","row_ids":["1fedb85017624b14b5d56927cae80633"],"score":59.26}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","peer_count":28,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6}],"score":59.26,"weight":10}],"eligible":true,"id":"m-8d99798a7e24690c","in_window":true,"name":"Gemini 3.5 Flash","official_evidence":{"artifact_sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","basis":"official_release_page","date":"2026-05-19","sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","url":"https://deepmind.google/models/model-cards/gemini-3-5-flash/"},"primary_profile_id":"native","profile_id":"native","rank":65,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","peer_count":28,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6}],"release_date":"2026-05-19","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":38.62,"score_exact":{"denominator":1116288,"numerator":43114225},"source_id":"google/gemini-3.5-flash"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":61047,"numerator":2006075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["04366109cc25918e59cfe4140983a31b"],"score":11.84},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2500},"family":"HLE","row_ids":["8319710b72494f93151d3337f0dbcdfd"],"score":39.68},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":800},"family":"CritPt","row_ids":["bd76f518c9968d380eda9d90a689c8c2"],"score":47.06}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","peer_count":64,"percentile":39.68,"percentile_exact":{"denominator":63,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","peer_count":52,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265}],"score":32.86,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":158004,"numerator":7953793},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1400},"family":"GDPval-AA","row_ids":["522de4d527da35b85f7ead36d5bab4a3"],"score":51.85},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":154},"family":"Terminal-Bench","row_ids":["7a6b6215e85870e72209d73fd4101db4"],"score":51.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1125},"family":"AutomationBench","row_ids":["981d2354449aefd49dcd7a2e96840b24"],"score":59.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3000},"family":"AA-Briefcase","row_ids":["fb238aac3df6d482092b3981302c93d1"],"score":38.96}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","peer_count":82,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":59.21,"percentile_exact":{"denominator":19,"numerator":1125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":78,"percentile":38.96,"percentile_exact":{"denominator":77,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2}],"score":50.34,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":91,"numerator":1825},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":200},"family":"FrontierCode","row_ids":["06196596577df812dcdeed161680d16c"],"score":28.57},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":150},"family":"Vals Code Migration","row_ids":["baba3c8fb65a67f1bc8fc2ae1572a4d3"],"score":11.54}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","peer_count":50,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","peer_count":27,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85}],"score":20.05,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":98025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1025},"family":"GDP.pdf","row_ids":["25ac15cd539fc43d4a096d3dea9f7966"],"score":53.95},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["11dc653c29e2dbe67eaa7a540042fef3","111f3e2d8b49ec87b3bbe30f22c6238c"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239}],"score":48.67,"weight":20}],"eligible":true,"id":"m-b6a574cc5300d5f1","in_window":true,"name":"Beam 1.0","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-16","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":66,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","peer_count":64,"percentile":39.68,"percentile_exact":{"denominator":63,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","peer_count":52,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","peer_count":82,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":59.21,"percentile_exact":{"denominator":19,"numerator":1125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":78,"percentile":38.96,"percentile_exact":{"denominator":77,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","peer_count":50,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","peer_count":27,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239}],"release_date":"2026-10-16","release_status":"published","release_verified":true,"result_updated_at":null,"score":38.38,"score_exact":{"denominator":137088952,"numerator":5261997115},"source_id":"published/aa:reflection-beam"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":5525},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":50},"family":"FrontierMath","row_ids":["34d11ef45eddcdc63e92f205eb9eaa3c"],"score":50.0},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":75},"family":"GPQA Diamond","row_ids":["6b87188fa1e95a23261b2a5e267b3413"],"score":37.5},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["9d8fe865edc8da70092f0b94926485a3"],"score":57.89}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0}],"score":48.46,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":12,"numerator":275},"families":[{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":175},"family":"Terminal-Bench","row_ids":["7fbb0050443eb76fe54b1727ad2dc1fe"],"score":43.75},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"AutomationBench","row_ids":["e4ce479712f2cad43715215ba9603696"],"score":25.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"AA-Briefcase","row_ids":["56c8ab5d498b1025e9684eb22709b6c7"],"score":0.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","peer_count":49,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334}],"score":22.92,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":75},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":75},"family":"DeepSWE","row_ids":["99e7d048583a82e4719e8e299872ac27"],"score":37.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6}],"score":37.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"GDP.pdf","row_ids":["2f6ca6bd54419e351a2ba1a4c64dedd7"],"score":50.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7}],"score":50.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["2d9aead127b28a667d7f4b6c145f9945"],"score":7.41},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["079f06724159c79071056bf02b730573"],"score":50.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","peer_count":28,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0}],"score":28.7,"weight":10}],"eligible":true,"id":"m-f8dbf8d16fd37aac","in_window":true,"name":"GPT-5.6 Terra","official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profile_id":"native","rank":67,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","peer_count":49,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","peer_count":28,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":"2026-07-09","score":38.22,"score_exact":{"denominator":8208,"numerator":313675},"source_id":"openai/gpt-5.6-terra"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1,"numerator":40},"families":[{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":40},"family":"GPQA Diamond","row_ids":["8afb3ab2b09d75bd6e3e04698e2c3cb8"],"score":40.0}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"reference_revision":"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4}],"score":40.0,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":18,"numerator":925},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["4ea5a3b77e03d03f1422a6728fc0742b"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":50},"family":"AA-Briefcase","row_ids":["680700051014bfbbbe521e958802f31c"],"score":5.56},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["0886e98ce0984c34f09c65e2184f8692"],"score":100.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":100},"family":"MCP Atlas","row_ids":["f29e77fd200946b25fe37a0df78120da"],"score":100.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"reference_revision":"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4}],"score":51.39,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["2206795ef677e458e6f834638793b28d"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"reference_revision":"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":34,"numerator":1945},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":70},"family":"IFBench","row_ids":["1be1bd504f0391002f7de4a46514d7c5"],"score":70.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"IFEval","row_ids":["98b4a1e95c1679056cdd34a216b7399d"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"MMLU-Pro","row_ids":["3e7111ab4e0b4463d98c531d78381779"],"score":100.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1000},"family":"SimpleQA Verified","row_ids":["cb8036f627af27545731c0fdb0405c49"],"score":58.82}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","peer_count":6,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"score":57.21,"weight":20}],"eligible":true,"id":"m-a3a34c5608c68d32","in_window":true,"name":"Qwen3.7 Max","official_evidence":{"artifact_sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","basis":"official_release_page","date":"2026-05-21","sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","url":"https://www.alibabacloud.com/help/en/model-studio/newly-released-models"},"primary_profile_id":"native","profile_id":"native","rank":68,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"reference_revision":"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"reference_revision":"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"reference_revision":"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","peer_count":6,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"release_date":"2026-05-21","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":38.1,"score_exact":{"denominator":5508,"numerator":209845},"source_id":"alibaba/qwen3.7-max"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":109575125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4100},"family":"GPQA Diamond","row_ids":["8e38fd38f9044aff514003f73b248716"],"score":69.49},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["a2c09e681871381cb89f80ae44384709"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":650},"family":"HLE","row_ids":["66e35510159ab65c6e04c83322148b65"],"score":10.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1850},"family":"CritPt","row_ids":["7de1f456199243fb6a03dcf75e8fd06c"],"score":36.27}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":69.49,"percentile_exact":{"denominator":59,"numerator":4100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18}],"score":33.35,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":58748329},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":725},"family":"tau3-Banking","row_ids":["87111691a6a525e89691d7ee783ff6e3"],"score":55.77},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3400},"family":"GDPval-AA","row_ids":["2235f722890b82a4475fc455a58c66ab"],"score":41.98},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["e5e132be2e915d3fe97356bbd06869f9"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":550},"family":"AutomationBench","row_ids":["ba2dbb617a983db0793fc0f1b006c6eb"],"score":28.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3900},"family":"AA-Briefcase","row_ids":["1caa339c8531983b8c2298b6a2b6d27c"],"score":50.65}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":55.77,"percentile_exact":{"denominator":13,"numerator":725},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","peer_count":82,"percentile":41.98,"percentile_exact":{"denominator":81,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":28.95,"percentile_exact":{"denominator":19,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":78,"percentile":50.65,"percentile_exact":{"denominator":77,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32}],"score":38.13,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":2014,"numerator":86325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":625},"family":"GDP.pdf","row_ids":["75cb1248936655503fa568c6e5bc1e02"],"score":32.89},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2800},"family":"AA-Omniscience","row_ids":["fcae50a41df6fb08e7a7f301689104ab","9364a9aa01cea849d2eb414760bb16ec"],"score":52.83}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255}],"score":42.86,"weight":20}],"eligible":true,"id":"m-754df8b5922bf5da","in_window":true,"name":"Solar Pro 4","official_evidence":{"artifact_sha256":"4c3c4258c97035773c61772893086ae86059b2a773780452e3c61615afc1a64f","basis":"official_release_page","date":"2026-08-11","sha256":"4c3c4258c97035773c61772893086ae86059b2a773780452e3c61615afc1a64f","url":"https://www.upstage.ai/blog/en/solar-pro-4"},"primary_profile_id":"native","profile_id":"native","rank":69,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":69.49,"percentile_exact":{"denominator":59,"numerator":4100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":55.77,"percentile_exact":{"denominator":13,"numerator":725},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","peer_count":82,"percentile":41.98,"percentile_exact":{"denominator":81,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":28.95,"percentile_exact":{"denominator":19,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":78,"percentile":50.65,"percentile_exact":{"denominator":77,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255}],"release_date":"2026-08-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":37.78,"score_exact":{"denominator":4586036711256,"numerator":173241935325145},"source_id":"upstage/solar-pro4"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":117650},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1300},"family":"GPQA Diamond","row_ids":["b5ca8a579b13fb2adc96c2c993971f7c"],"score":22.03},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2600},"family":"HLE","row_ids":["92b0eaaad67a910391a173c1c1816701"],"score":41.27}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":22.03,"percentile_exact":{"denominator":59,"numerator":1300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","peer_count":64,"percentile":41.27,"percentile_exact":{"denominator":63,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131}],"score":31.65,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":391169138},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":825},"family":"tau3-Banking","row_ids":["9ed3eb345da97e474521a1cd387dd7bc"],"score":63.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4300},"family":"GDPval-AA","row_ids":["95847d736743738d4e269c199a41aca9"],"score":53.09},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["84fdb274483242a5de60a1387cb766c1"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":825},"family":"AutomationBench","row_ids":["94674bbf056258753d15a53bade2e692"],"score":43.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4000},"family":"AA-Briefcase","row_ids":["f0265d9e7d722340c4dcae3d534155b9"],"score":51.95}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":63.46,"percentile_exact":{"denominator":13,"numerator":825},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","peer_count":82,"percentile":53.09,"percentile_exact":{"denominator":81,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":43.42,"percentile_exact":{"denominator":19,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","peer_count":78,"percentile":51.95,"percentile_exact":{"denominator":77,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55}],"score":50.78,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1075},"family":"GDP.pdf","row_ids":["3fef1cbee3ec381cc5e40ecf79f229ab"],"score":28.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":28.29,"weight":20}],"eligible":true,"id":"m-71070d6294f8dc57","in_window":true,"name":"Quasar 438B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":70,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":22.03,"percentile_exact":{"denominator":59,"numerator":1300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","peer_count":64,"percentile":41.27,"percentile_exact":{"denominator":63,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":63.46,"percentile_exact":{"denominator":13,"numerator":825},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","peer_count":82,"percentile":53.09,"percentile_exact":{"denominator":81,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":43.42,"percentile_exact":{"denominator":19,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","peer_count":78,"percentile":51.95,"percentile_exact":{"denominator":77,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"release_date":"2026-08-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":37.52,"score_exact":{"denominator":636242607,"numerator":23874290621},"source_id":"published/aa:quasar-438b"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["6d2464487e35f420dae74891c6e0dec0"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":513513,"numerator":11596703},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":400},"family":"tau3-Banking","row_ids":["f6ef999603d2cd2e7f3ecd5e4118aa3e"],"score":30.77},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":800},"family":"GDPval-AA","row_ids":["518222bbae6a6938c2fc68e86213e044"],"score":29.63},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["46a72c291abf86482bd30a1de9ec49ac"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":325},"family":"AutomationBench","row_ids":["3e7bd955b2670c77e55665a0183fa94c"],"score":17.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1700},"family":"AA-Briefcase","row_ids":["628eb864f68ddd810e173dc996c88c01"],"score":22.08}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","peer_count":82,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":17.11,"percentile_exact":{"denominator":19,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","peer_count":78,"percentile":22.08,"percentile_exact":{"denominator":77,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17}],"score":22.58,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1175},"family":"GDP.pdf","row_ids":["10af516d49f62278a451fea09c7878c1"],"score":30.92}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"score":30.92,"weight":20}],"eligible":true,"id":"m-d48cde87e1f9c28d","in_window":true,"name":"A.X-K2","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":71,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","peer_count":82,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":17.11,"percentile_exact":{"denominator":19,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","peer_count":78,"percentile":22.08,"percentile_exact":{"denominator":77,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":36.88,"score_exact":{"denominator":424161738,"numerator":15641285435},"source_id":"published/aa:a-x-k2"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":57,"numerator":1375},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":50},"family":"GPQA Diamond","row_ids":["021a82f4f7be85e4b693852ce483ef03"],"score":16.67},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["5a0b6bca7e87cdf2b14ffe16a27cf187"],"score":31.58}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":24.12,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":54,"numerator":1975},"families":[{"benchmark_version":"4.0","exact":{"denominator":2,"numerator":75},"family":"Terminal-Bench","row_ids":["6bb0c331de392cf1dc35a1de5de30317"],"score":37.5},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["25c531e52c90edae0ad06f8953351733"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":650},"family":"AA-Briefcase","row_ids":["239fbb8019bce72c55d5b79e4b9186e5"],"score":72.22}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","peer_count":49,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516}],"score":36.57,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3,"numerator":200},"families":[{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":200},"family":"DeepSWE","row_ids":["72c84440fb70eb9d01c3241fc4d3e551"],"score":66.67}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4}],"score":66.67,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"IFBench","row_ids":["bb7586265eb9f4ee440c1de9eb72708e"],"score":33.33},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":200},"family":"SimpleQA Verified","row_ids":["f310a7a1d6e944a484a86a2598b55b31"],"score":11.76}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636}],"score":22.55,"weight":20}],"eligible":true,"id":"m-3535a3e7891af5ef","in_window":true,"name":"DeepSeek V4 Flash 0731","official_evidence":{"artifact_sha256":"3f80af72789dd0dd52b0bd5de471e2df2bc33c6eb443bf98eaeecc4dec6afbf4","basis":"official_release_page","date":"2026-07-31","sha256":"3f80af72789dd0dd52b0bd5de471e2df2bc33c6eb443bf98eaeecc4dec6afbf4","url":"https://api-docs.deepseek.com/updates/"},"primary_profile_id":"native","profile_id":"native","rank":72,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","peer_count":49,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636}],"release_date":"2026-07-31","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":36.69,"score_exact":{"denominator":313956,"numerator":11517775},"source_id":"deepseek/deepseek-v4-flash-0731"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":413,"numerator":23125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4500},"family":"GPQA Diamond","row_ids":["9b6a2e4e945c90d8754fc862b2e6383b"],"score":76.27},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":250},"family":"HLE","row_ids":["a0be97a286389603a66382ac4d794e81"],"score":35.71}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":76.27,"percentile_exact":{"denominator":59,"numerator":4500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116}],"score":55.99,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":1320462,"numerator":43715939},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["869a19a4a7bb34031859441309a0c7b5"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":700},"family":"tau3-Banking","row_ids":["a00dadd2de4cef1d6ffa67d4028686b0"],"score":53.85},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3100},"family":"GDPval-AA","row_ids":["3e2d18ecd2bf6e50f2e15788da94966d"],"score":38.27},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["87cb9fb97ee59be625938b752ce0bb36"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":725},"family":"AutomationBench","row_ids":["df8bf63c05795b45b4fd82cd57c8b504"],"score":38.16},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":400},"family":"AA-Briefcase","row_ids":["a684e0a09f0e331ba14d29191ea44f68"],"score":36.36}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","peer_count":82,"percentile":38.27,"percentile_exact":{"denominator":81,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","peer_count":78,"percentile":36.36,"percentile_exact":{"denominator":11,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24}],"score":33.11,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["6f7ac5ece00998e78042608639ac6e80"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["2081b40577b770d6210bb3099918429f"],"score":48.03}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"score":48.03,"weight":20}],"eligible":true,"id":"m-d23e83fbb1d3aabc","in_window":true,"name":"Hy3","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":73,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":76.27,"percentile_exact":{"denominator":59,"numerator":4500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","peer_count":82,"percentile":38.27,"percentile_exact":{"denominator":81,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","peer_count":78,"percentile":36.36,"percentile_exact":{"denominator":11,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"release_date":"2026-07-06","release_status":"published","release_verified":true,"result_updated_at":null,"score":35.42,"score_exact":{"denominator":9816314508,"numerator":347716592885},"source_id":"tencent/hy3"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":1850},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1850},"family":"GPQA Diamond","row_ids":["6fa185e6f3a52003678d56e35ea14de0"],"score":31.36}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252}],"score":31.36,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1539,"numerator":10700},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":700},"family":"GDPval-AA","row_ids":["60ed1cde923cee15c4935672b177aa62"],"score":8.64},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":100},"family":"AutomationBench","row_ids":["49cf1a20c5edfdf8e5e8bfd178db37ad"],"score":5.26}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","peer_count":82,"percentile":8.64,"percentile_exact":{"denominator":81,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258}],"score":6.95,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4,"numerator":325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":325},"family":"IFBench","row_ids":["af2b7a8fe076debb535a31ef7b09244c"],"score":81.25}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864}],"score":81.25,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":250},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":250},"family":"MMMU-Pro","row_ids":["ca0bc4ca214d2607b60ef7f92f73e72e"],"score":14.71}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514}],"score":14.71,"weight":10}],"eligible":true,"id":"m-a4be17f49ae915ce","in_window":true,"name":"Gemma 4 12B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":74,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","peer_count":82,"percentile":8.64,"percentile_exact":{"denominator":81,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514}],"release_date":"2026-06-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":34.12,"score_exact":{"denominator":24697872,"numerator":842744275},"source_id":"published/aa:gemma-4-12b"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["7c25d00a19511ccc369639cf5605291e"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":183027373},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":500},"family":"tau3-Banking","row_ids":["3cd21f0f8001e029d7b19208e29ba099"],"score":38.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1700},"family":"GDPval-AA","row_ids":["2a88f9d5ec82d3f04004cf3da53914a5"],"score":20.99},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["7e0782bec5efa3fc907fbba656122a79"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":125},"family":"AutomationBench","row_ids":["bfaca7bcdefcec6c5d1daac93dc0d456"],"score":6.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1600},"family":"AA-Briefcase","row_ids":["d8f9d30bc51b5ac382358dbda4a533e5"],"score":20.78}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","peer_count":82,"percentile":20.99,"percentile_exact":{"denominator":81,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":78,"percentile":20.78,"percentile_exact":{"denominator":77,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95}],"score":23.76,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":12084,"numerator":238625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":75},"family":"IFBench","row_ids":["3080cdd2a4ddf6bfb1a2fed6805cd96e"],"score":18.75},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":375},"family":"GDP.pdf","row_ids":["5ecbc58b2efe8e4c7ad00b00fd409205"],"score":19.74},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1100},"family":"AA-Omniscience","row_ids":["f5ca62e67f09a75d5d460c81d0851fa4","2c99a44c194b274fedb14bdd3896ace9"],"score":20.75}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393}],"score":19.75,"weight":20}],"eligible":true,"id":"m-25ccaf97cf82a5e0","in_window":true,"name":"Ring-2.6-1T","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-08","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":75,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","peer_count":82,"percentile":20.99,"percentile_exact":{"denominator":81,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":78,"percentile":20.78,"percentile_exact":{"denominator":77,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393}],"release_date":"2026-05-08","release_status":"published","release_verified":true,"result_updated_at":null,"score":34.1,"score_exact":{"denominator":33720858171,"numerator":1150018852748},"source_id":"published/aa:ring-2-6-1t"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":7182,"numerator":275461},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1100},"family":"GDPval-AA","row_ids":["dde8616d382b2b0a447eb1104181da5e"],"score":40.74},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["7620c0fb57936004d0ff80edb6675f9f"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":800},"family":"AutomationBench","row_ids":["b1f8ac835c2c56f569ae7093d49e508f"],"score":42.11},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":200},"family":"AA-Briefcase","row_ids":["b76daa2b8d1a11ba3f7dd4198d33e0a0"],"score":28.57}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","peer_count":82,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":42.11,"percentile_exact":{"denominator":19,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":78,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74}],"score":38.35,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":4028,"numerator":113625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":925},"family":"GDP.pdf","row_ids":["d7e2c6c9bb5b766f6065fc5409408479"],"score":24.34},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1700},"family":"AA-Omniscience","row_ids":["8e92630a337eaf823218ceb658e6daf8","973d182535a01b720897f5a733a8d1b0"],"score":32.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599}],"score":28.21,"weight":20}],"eligible":true,"id":"m-f99cd5ceed3a5b41","in_window":true,"name":"Solar Mini 4","official_evidence":{"artifact_sha256":"967417099d792a2cbc0abac952f572bc529067f8ba0587e3c041a5f253bc2a0e","basis":"official_release_page","date":"2026-10-01","sha256":"967417099d792a2cbc0abac952f572bc529067f8ba0587e3c041a5f253bc2a0e","url":"https://www.upstage.ai/blog/en/solar-mini-4"},"primary_profile_id":"native","profile_id":"native","rank":76,"records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","peer_count":82,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":42.11,"percentile_exact":{"denominator":19,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":78,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599}],"release_date":"2026-10-01","release_status":"published","release_verified":true,"result_updated_at":null,"score":33.85,"score_exact":{"denominator":3425814,"numerator":115947415},"source_id":"upstage/solar-mini4"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":850},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["01d54ae905ebb77f621f6cbf30eb0930"],"score":44.74}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"score":44.74,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":96,"numerator":2525},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":100},"family":"GDPval-AA","row_ids":["fcb5b9889977ea6f8824ae1c412b655a"],"score":11.11},{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":175},"family":"Terminal-Bench","row_ids":["bbf181bad55ef3d8ecad294914ed6b67"],"score":21.88},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":500},"family":"AA-Briefcase","row_ids":["14b2d91b066241cbb4ba0914b53ae0a4"],"score":55.56},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":50},"family":"OSWorld","row_ids":["cdb864bc0c75bd1fb1f8c43c173d544c"],"score":16.67}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"v1.1","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0}],"score":26.3,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["494d4b4ce09501b6feb0dbb80233b707"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":150},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":150},"family":"SimpleQA Verified","row_ids":["0c6f13f860966ddef45cc04368fd5127"],"score":8.82}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329}],"score":8.82,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":250},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":500},"family":"Humanity Sixth Sense","row_ids":["cf4a733238c901c9de673ff40841fce0"],"score":18.52},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["a4e7fdc6e1f2295b627a11265f28666d"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","peer_count":28,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6}],"score":9.26,"weight":10}],"eligible":true,"id":"m-2cd37409a483435a","in_window":true,"name":"Claude Sonnet 5","official_evidence":{"artifact_sha256":"78c9b451825b0f421c2e785a5934defd9662507c3913b2f7c2c0df07917ba424","basis":"official_release_page","date":"2026-06-30","sha256":"78c9b451825b0f421c2e785a5934defd9662507c3913b2f7c2c0df07917ba424","url":"https://www.anthropic.com/news/claude-sonnet-5"},"primary_profile_id":"native","profile_id":"native","rank":77,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"v1.1","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","peer_count":28,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6}],"release_date":"2026-06-30","release_status":"published","release_verified":true,"result_updated_at":"2026-09-28","score":32.95,"score_exact":{"denominator":1116288,"numerator":36782095},"source_id":"anthropic/claude-sonnet-5"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":41769,"numerator":1004075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["42505f6b2ea9972da7c497ad966aebac"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1600},"family":"HLE","row_ids":["30beb0844cc09771243cbe9dfff243cd"],"score":25.4},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":500},"family":"CritPt","row_ids":["e6e2851045fbef435d3da74124d06523"],"score":29.41}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"Diamond","peer_count":64,"percentile":25.4,"percentile_exact":{"denominator":63,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","peer_count":52,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135}],"score":24.04,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":10773,"numerator":369385},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4000},"family":"GDPval-AA","row_ids":["3c9b5c57e54a04df82602451589cb092"],"score":49.38},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["01e4bcd3555fdf1970b9b729ff28db44"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":600},"family":"AutomationBench","row_ids":["ae26dd2850d89ddaa6b644b249715c9f"],"score":31.58},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":300},"family":"AA-Briefcase","row_ids":["b0a10dcacf57ed9fc64b8506d23f5ae1"],"score":42.86}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2.1","peer_count":82,"percentile":49.38,"percentile_exact":{"denominator":81,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":78,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76}],"score":34.29,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1600},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1600},"family":"FrontierCode","row_ids":["35228bba34f7d198ea6ae7b85139d6dc"],"score":32.65}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"2.0","peer_count":50,"percentile":32.65,"percentile_exact":{"denominator":49,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141}],"score":32.65,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":79475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":675},"family":"GDP.pdf","row_ids":["26277d899631407b16323e0f427ca69b"],"score":35.53},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["41481c594a0d6acdcc3cd6076cd49938","1c32657b7924b6111d38086436775d4a"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161}],"score":39.46,"weight":20}],"eligible":true,"id":"m-36451e3f44503d37","in_window":true,"name":"Nemotron 3.5 Super VL","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":78,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"Diamond","peer_count":64,"percentile":25.4,"percentile_exact":{"denominator":63,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","peer_count":52,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2.1","peer_count":82,"percentile":49.38,"percentile_exact":{"denominator":81,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":78,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"2.0","peer_count":50,"percentile":32.65,"percentile_exact":{"denominator":49,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"result_updated_at":null,"score":32.23,"score_exact":{"denominator":7949601387,"numerator":256194148075},"source_id":"published/aa:nemotron-3-5-super-vl"},{"benchmark_count":20,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":26078,"numerator":848075},"families":[{"benchmark_version":"2026","exact":{"denominator":3,"numerator":200},"family":"AIME","row_ids":["9bcefe5681f1de0d56a93cb5ca7a6411"],"score":66.67},{"benchmark_version":"500-item subset","exact":{"denominator":1,"numerator":100},"family":"MATH-500","row_ids":["48e7cd14f60fe5daf4244ca5ca1c19b2"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":600},"family":"GPQA Diamond","row_ids":["7e93a07b82dfd690e1c2b1309992332e"],"score":10.17},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d1c19809bb6f7e59cd9274e82fb1b4b5"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":0},"family":"HLE","row_ids":["20cab07f9ffc43544259f41d417e323c"],"score":0.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":50},"family":"CritPt","row_ids":["e23c22def1e7c8613f7559601fcfa581"],"score":0.98}],"family_count":6,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":10.17,"percentile_exact":{"denominator":59,"numerator":600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","peer_count":64,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":32.52,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":6162156,"numerator":60705805},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":125},"family":"tau3-Banking","row_ids":["9e32e2abcb9a8a594fb5e6aa3bcbd501"],"score":4.81},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":100},"family":"GDPval-AA","row_ids":["a9600d8248b3ea039e7316ffc060f044"],"score":3.7},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["c56ed05cd995d4a9a967c8a8edd7c23f"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":50},"family":"AutomationBench","row_ids":["de96f31cb9e7df111014b360185583e3"],"score":2.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":100},"family":"AA-Briefcase","row_ids":["caf840132854029b414465a52c19889d"],"score":1.3},{"benchmark_version":"4","exact":{"denominator":3,"numerator":100},"family":"BFCL","row_ids":["c9019a0cb657ee535803619a061368cf"],"score":33.33}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","peer_count":82,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":2.63,"percentile_exact":{"denominator":19,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":78,"percentile":1.3,"percentile_exact":{"denominator":77,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2}],"score":9.85,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":147,"numerator":5125},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["d294135a5f841df5749d74dee1cdb9e7"],"score":3.06},{"benchmark_version":"release_v6","exact":{"denominator":3,"numerator":200},"family":"LiveCodeBench","row_ids":["2f001b0be2d50fef9306d2b3f9f1569b"],"score":66.67}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9}],"score":34.86,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":477,"numerator":25225},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"IFBench","row_ids":["a7d850e0b3d6527713784858f859aa3f"],"score":66.67},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":200},"family":"Multi-IF","row_ids":["cf2e1f5dc39c9244f47a835c7f33569d"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"GDP.pdf","row_ids":["d9c7d28e2d9651e546a84191d6ab8ec7"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"IFEval","row_ids":["55369a687291e4012c741d522a892ee5"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":50},"family":"MMLU-Pro","row_ids":["8519642492aa86c8bdd1ad71a0f84568"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["432ed3bb705952b8c4dcd7c697936b37","7f11b1b4129f0979e12158aab3657dcc"],"score":33.96}],"family_count":6,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509}],"score":52.88,"weight":20}],"eligible":true,"id":"m-12d2cbffeaae72d5","in_window":true,"name":"Granite 4.2 3B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":79,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":10.17,"percentile_exact":{"denominator":59,"numerator":600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","peer_count":64,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","peer_count":82,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":2.63,"percentile_exact":{"denominator":19,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":78,"percentile":1.3,"percentile_exact":{"denominator":77,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.27,"score_exact":{"denominator":41274330401304,"numerator":1290617742549175},"source_id":"published/aa:granite-4-2-3b"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":117351,"numerator":3265325},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2400},"family":"GPQA Diamond","row_ids":["da90bd5db2e44d989ddf563c8ec38c6b"],"score":40.68},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["13146daedf685d8da2f4bf015457ee2c"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1300},"family":"CritPt","row_ids":["d6e7cb95527cd29a8925ca94f594e568"],"score":25.49}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":40.68,"percentile_exact":{"denominator":59,"numerator":2400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","peer_count":52,"percentile":25.49,"percentile_exact":{"denominator":51,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08}],"score":27.83,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":24453,"numerator":1164503},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":875},"family":"tau3-Banking","row_ids":["7e3c134e84deebeefe2308a65aef1578"],"score":67.31},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":500},"family":"GDPval-AA","row_ids":["7f3c9ed4efe6520cbbf9a35bd5a39f5e"],"score":55.56},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["cd7e16f694d6c0543ca83b3b5dd9358b"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":900},"family":"AutomationBench","row_ids":["908bb98659898bf60c101cff97ce69b8"],"score":47.37},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":600},"family":"AA-Briefcase","row_ids":["54e0dbe28b524361d2c39e892adaa185"],"score":54.55}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":67.31,"percentile_exact":{"denominator":13,"numerator":875},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","peer_count":82,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":47.37,"percentile_exact":{"denominator":19,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","peer_count":78,"percentile":54.55,"percentile_exact":{"denominator":11,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79}],"score":47.62,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1200},"family":"FrontierCode","row_ids":["33a3261e51180ab895a82ccd6f7e12a3"],"score":24.49}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","peer_count":50,"percentile":24.49,"percentile_exact":{"denominator":49,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444}],"score":24.49,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":825},"family":"GDP.pdf","row_ids":["14efb36fead599ff0b956fa101f23a86"],"score":21.71}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042}],"score":21.71,"weight":20}],"eligible":true,"id":"m-a67d9ce95a33f246","in_window":true,"name":"K2 Horizon MoVA 36B A4B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":80,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":40.68,"percentile_exact":{"denominator":59,"numerator":2400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","peer_count":52,"percentile":25.49,"percentile_exact":{"denominator":51,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":67.31,"percentile_exact":{"denominator":13,"numerator":875},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","peer_count":82,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":47.37,"percentile_exact":{"denominator":19,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","peer_count":78,"percentile":54.55,"percentile_exact":{"denominator":11,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","peer_count":50,"percentile":24.49,"percentile_exact":{"denominator":49,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.22,"score_exact":{"denominator":3605374773,"numerator":112575457730},"source_id":"published/aa:k2-horizon-mova-36b-a4b"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":1400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1400},"family":"GPQA Diamond","row_ids":["bbee1bae6aa788cf47294ddc38f9761f"],"score":23.73}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":23.73,"percentile_exact":{"denominator":59,"numerator":1400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333}],"score":23.73,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":648648,"numerator":5385785},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":125},"family":"tau3-Banking","row_ids":["24158d9869238dabbd834ff7fd267dc4"],"score":4.81},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":800},"family":"GDPval-AA","row_ids":["fc44351a5b9d8e210a763f56506c0719"],"score":9.88},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["2b6eac192b734b825e56a7c642b341b7"],"score":13.33},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":400},"family":"AA-Briefcase","row_ids":["606c6cc32d7931469129472a4d6fffcd"],"score":5.19}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","peer_count":82,"percentile":9.88,"percentile_exact":{"denominator":81,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","peer_count":78,"percentile":5.19,"percentile_exact":{"denominator":77,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41}],"score":8.3,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4,"numerator":275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":275},"family":"IFBench","row_ids":["ea00f82eb447a348674a4f2b40a4a4f4"],"score":68.75}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136}],"score":68.75,"weight":20}],"eligible":true,"id":"m-7b2a8fe101678ebd","in_window":true,"name":"HyperNova 60B 2605","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":81,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":23.73,"percentile_exact":{"denominator":59,"numerator":1400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","peer_count":82,"percentile":9.88,"percentile_exact":{"denominator":81,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","peer_count":78,"percentile":5.19,"percentile_exact":{"denominator":77,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136}],"release_date":"2026-05-26","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.08,"score_exact":{"denominator":535783248,"numerator":16653656375},"source_id":"published/aa:hypernova-60b"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1095276,"numerator":37020775},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3700},"family":"GPQA Diamond","row_ids":["198f62e91fe9f1a3e9e87c6a5a19a9fe"],"score":62.71},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["70737e236c952ff057602f3190f37b75"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":500},"family":"HLE","row_ids":["85bd1df1c636ca5159d1bed6e1a0f499"],"score":23.81},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1600},"family":"CritPt","row_ids":["f5fbcfce1c679af1818579e2368d3d27"],"score":31.37}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":62.71,"percentile_exact":{"denominator":59,"numerator":3700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","peer_count":64,"percentile":23.81,"percentile_exact":{"denominator":21,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","peer_count":52,"percentile":31.37,"percentile_exact":{"denominator":51,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15}],"score":33.8,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":10783773,"numerator":221054348},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":300},"family":"tau3-Banking","row_ids":["5c824422e02d6ae8e8e0e7805866dd38"],"score":23.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2800},"family":"GDPval-AA","row_ids":["d9fac2d0e8fa2fae504adca322b42dfa"],"score":34.57},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["22a0f772949251594cd3a25e02ef9752"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":350},"family":"AutomationBench","row_ids":["b55d26365439e3d2ed6eba908f339027"],"score":18.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2300},"family":"AA-Briefcase","row_ids":["e12061e69225a05e579dff8f6ab0f0de"],"score":29.87},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["4c1c120149f7031df34c6f3ac8ff9629"],"score":5.56},{"benchmark_version":"current AA pass^5","exact":{"denominator":1,"numerator":0},"family":"AA-AnalystAgent","row_ids":["2275301a736d0fe2b34ce5c990945f54"],"score":0.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","peer_count":82,"percentile":34.57,"percentile_exact":{"denominator":81,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":78,"percentile":29.87,"percentile_exact":{"denominator":77,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","peer_count":24,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625}],"score":20.5,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":7,"numerator":100},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":100},"family":"FrontierCode","row_ids":["07260be02cf7c09201c1af77d371a386"],"score":14.29}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","peer_count":50,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576}],"score":14.29,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":232025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":375},"family":"IFBench","row_ids":["74e2baf55ca27065ec7aeb8f6eee9414"],"score":93.75},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":925},"family":"GDP.pdf","row_ids":["994833badab5af26add99f20244353b3"],"score":24.34},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2900},"family":"AA-Omniscience","row_ids":["7a931d4238635511e69e1feb35bfd055","f659a47d7f8491b1073ef450bc951aa0"],"score":54.72}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","peer_count":17,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645}],"score":57.6,"weight":20}],"eligible":true,"id":"m-ad7decc9ae94df12","in_window":true,"name":"Nemotron 3 Ultra 550B A55B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-04","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":82,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":62.71,"percentile_exact":{"denominator":59,"numerator":3700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","peer_count":64,"percentile":23.81,"percentile_exact":{"denominator":21,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","peer_count":52,"percentile":31.37,"percentile_exact":{"denominator":51,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","peer_count":82,"percentile":34.57,"percentile_exact":{"denominator":81,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":78,"percentile":29.87,"percentile_exact":{"denominator":77,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","peer_count":24,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","peer_count":50,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","peer_count":17,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645}],"release_date":"2026-06-04","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.06,"score_exact":{"denominator":41274330401304,"numerator":1281914164344215},"source_id":"nvidia/nemotron-3-ultra-550b-a55b"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":177,"numerator":500},"families":[{"benchmark_version":"2026","exact":{"denominator":1,"numerator":0},"family":"AIME","row_ids":["1aa7620d5715919ff1d95bb835bf5bb1"],"score":0.0},{"benchmark_version":"500-item subset","exact":{"denominator":1,"numerator":0},"family":"MATH-500","row_ids":["d2a6fbf18a7d28ad26aab8f6363bd884"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":500},"family":"GPQA Diamond","row_ids":["c0382c8dd0c506272b3fef2eaf1a7bcf"],"score":8.47}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":8.47,"percentile_exact":{"denominator":59,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758}],"score":2.82,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3159,"numerator":120050},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":150},"family":"tau3-Banking","row_ids":["ad5761dce64ad5d75bd8fe071f2c9120"],"score":11.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":200},"family":"GDPval-AA","row_ids":["7d56dec4d055d3d9b1237e11f0bf7ef4"],"score":2.47},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["78759a40401be7663e3957b80c649826"],"score":100.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","peer_count":82,"percentile":2.47,"percentile_exact":{"denominator":81,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88}],"score":38.0,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"release_v6","exact":{"denominator":3,"numerator":100},"family":"LiveCodeBench","row_ids":["ff54ed0b44635ad5d8a708f3d17d4063"],"score":33.33}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1}],"score":33.33,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":36,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"IFBench","row_ids":["49b7d39b3c4a3eaca06d3883d0335282"],"score":11.11},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"Multi-IF","row_ids":["ddcaace3bb0b4a1452f5dea314dd058c"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":200},"family":"IFEval","row_ids":["a24e132d4558f677a1ce0a399471dbb9"],"score":66.67},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":25},"family":"MMLU-Pro","row_ids":["239f2b6a72085e3eb63d599375b5b516"],"score":25.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2}],"score":50.69,"weight":20}],"eligible":true,"id":"m-d0874ed91fd69e8e","in_window":true,"name":"LFM2.5-2.6B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-04","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":83,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":8.47,"percentile_exact":{"denominator":59,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","peer_count":82,"percentile":2.47,"percentile_exact":{"denominator":81,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2}],"release_date":"2026-08-04","release_status":"published","release_verified":true,"result_updated_at":null,"score":30.01,"score_exact":{"denominator":3354858,"numerator":100691975},"source_id":"published/aa:lfm2-5-2-6b"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":123000875},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3800},"family":"GPQA Diamond","row_ids":["a95cdd62235bf41ff0b78b6a5a866d88"],"score":64.41},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["8120a268fd2b22fe2a65df299e1ab0c5"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2000},"family":"HLE","row_ids":["2d291e3811085b2648de857339509203"],"score":31.75},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1850},"family":"CritPt","row_ids":["a9467dab432c3fd1ab9b333380c66f83"],"score":36.27}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":64.41,"percentile_exact":{"denominator":59,"numerator":3800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","included":true,"model_id":"m-b439c804161fac26","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond","peer_count":64,"percentile":31.75,"percentile_exact":{"denominator":63,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18}],"score":37.43,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":70864794,"numerator":2278099781},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":850},"family":"tau3-Banking","row_ids":["22b4395fa3e0e77e06e9fcfe630f8789"],"score":65.38},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3200},"family":"GDPval-AA","row_ids":["ff011af995e95f243ad573421cc4ba05"],"score":39.51},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["1344e9967afa11034830d610f592ed83"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":450},"family":"AutomationBench","row_ids":["5bd1e6ef9a7eb6f30e6c6bf694074f2d"],"score":23.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2000},"family":"AA-Briefcase","row_ids":["7d5fcb038f359de5556a57151b51cffc"],"score":25.97},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":350},"family":"EnterpriseOps-Gym","row_ids":["40747d0e01fbdeb26f3ad425860f13bb"],"score":38.89},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["73ec2c70e1808cc25e2b747786aa5d99"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":500},"family":"AA-AnalystAgent","row_ids":["753dbaee803264fefa51f263a84b1509"],"score":21.74}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","included":true,"model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2.1","peer_count":82,"percentile":39.51,"percentile_exact":{"denominator":81,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":23.68,"percentile_exact":{"denominator":19,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":78,"percentile":25.97,"percentile_exact":{"denominator":77,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA run","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","peer_count":24,"percentile":21.74,"percentile_exact":{"denominator":23,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375}],"score":32.15,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":49,"numerator":550},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1100},"family":"FrontierCode","row_ids":["9b48c9ab902cc3f53ca099a627db7e76"],"score":22.45},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"Vals Code Migration","row_ids":["2bad0033060ca0f870378d7a7d3d6c5d"],"score":0.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","included":true,"model_id":"m-b439c804161fac26","original_version_label":"2.0","peer_count":50,"percentile":22.45,"percentile_exact":{"denominator":49,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1.1","peer_count":27,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82}],"score":11.22,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2115350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1150},"family":"GDP.pdf","row_ids":["af7da6ed00064e0d491020c30636379c"],"score":60.53},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":350},"family":"SimpleQA Verified","row_ids":["28411df2fb9e1660a30498ca8b0c9245"],"score":20.59},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["95037c24ac83fd0eadfbf39605ac569f","3698e24fd77dd4f8bc27edd44ed2efc7"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","included":true,"model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329}],"score":41.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":350},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":350},"family":"MMMU-Pro","row_ids":["874b32395a8c924c1d3f0faf3d2e2b2e"],"score":20.59}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856}],"score":20.59,"weight":10}],"eligible":true,"id":"m-b439c804161fac26","in_window":true,"name":"Inkling","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":84,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":64.41,"percentile_exact":{"denominator":59,"numerator":3800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","included":true,"model_id":"m-b439c804161fac26","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond","peer_count":64,"percentile":31.75,"percentile_exact":{"denominator":63,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","included":true,"model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2.1","peer_count":82,"percentile":39.51,"percentile_exact":{"denominator":81,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":23.68,"percentile_exact":{"denominator":19,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":78,"percentile":25.97,"percentile_exact":{"denominator":77,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA run","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","peer_count":24,"percentile":21.74,"percentile_exact":{"denominator":23,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","included":true,"model_id":"m-b439c804161fac26","original_version_label":"2.0","peer_count":50,"percentile":22.45,"percentile_exact":{"denominator":49,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1.1","peer_count":27,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","included":true,"model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856}],"release_date":"2026-07-15","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":29.94,"score_exact":{"denominator":16227514516752,"numerator":485799335327597},"source_id":"thinkingmachines/inkling"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":365092,"numerator":15288225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4300},"family":"GPQA Diamond","row_ids":["70d4757947a89f4f143b905fc208d5b4"],"score":72.88},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["569c655345f8f6c2c6fa591358c80baa"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":800},"family":"HLE","row_ids":["3c34699bd1ded8eb4f916044a6a266ef"],"score":38.1},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2000},"family":"CritPt","row_ids":["0bfa1e021f0f3d031ecf9fe9a047c9e1"],"score":39.22}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":72.88,"percentile_exact":{"denominator":59,"numerator":4300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","peer_count":64,"percentile":38.1,"percentile_exact":{"denominator":21,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","peer_count":52,"percentile":39.22,"percentile_exact":{"denominator":51,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185}],"score":41.87,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":755891136,"numerator":23344506125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":525},"family":"tau3-Banking","row_ids":["b9bbbaa648034918f0dabaffc06da3bd"],"score":40.38},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1300},"family":"GDPval-AA","row_ids":["fd7e6b8b2243df420c1327f60f09d326"],"score":48.15},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":25},"family":"Terminal-Bench","row_ids":["2f300304d9d00cf74194cf028a1e2a05"],"score":1.04},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":425},"family":"AutomationBench","row_ids":["441205ca0d2d0b6e786d210564b45527"],"score":22.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2500},"family":"AA-Briefcase","row_ids":["b87f8c47a158b30537b60e5453c5b4d9"],"score":32.47},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":550},"family":"EnterpriseOps-Gym","row_ids":["fe3cd8a1335112d2ed6891e37698119e"],"score":61.11},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":100},"family":"MCP Atlas","row_ids":["72495a708e9f00af5b2d17370ac07628"],"score":11.11},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":700},"family":"AA-AnalystAgent","row_ids":["6890e628f5988bd5a798a680de5a69d4"],"score":30.43}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":40.38,"percentile_exact":{"denominator":13,"numerator":525},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","peer_count":82,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"4.0","peer_count":49,"percentile":1.04,"percentile_exact":{"denominator":24,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":22.37,"percentile_exact":{"denominator":19,"numerator":425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":78,"percentile":32.47,"percentile_exact":{"denominator":77,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","peer_count":24,"percentile":30.43,"percentile_exact":{"denominator":23,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275}],"score":30.88,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":9675},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1300},"family":"FrontierCode","row_ids":["066fdac9a6f7fc061fce980c5a691276"],"score":26.53},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":50},"family":"Vals Code Migration","row_ids":["17347953383a652f8a8fabeabf1d07e6"],"score":3.85}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"2.0","peer_count":50,"percentile":26.53,"percentile_exact":{"denominator":49,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1.1","peer_count":27,"percentile":3.85,"percentile_exact":{"denominator":13,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43}],"score":15.19,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":6042,"numerator":170825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["17242a9341b3cc9173ea459907ce9455"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":0},"family":"SimpleQA Verified","row_ids":["62e95547f0f7073efcdd1420b882e4a5"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1950},"family":"AA-Omniscience","row_ids":["c31ed6f5932783f10dbc8e7e910353cc","01245c922f864bcfcfa4ad1336dea016"],"score":36.79}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433}],"score":28.27,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":400},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":400},"family":"MMMU-Pro","row_ids":["93ce1046100f7e346ed0a342303a4604"],"score":23.53}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665}],"score":23.53,"weight":10}],"eligible":true,"id":"m-35c4f287acc76c95","in_window":true,"name":"Inkling Small","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-30","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":85,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":72.88,"percentile_exact":{"denominator":59,"numerator":4300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","peer_count":64,"percentile":38.1,"percentile_exact":{"denominator":21,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","peer_count":52,"percentile":39.22,"percentile_exact":{"denominator":51,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":40.38,"percentile_exact":{"denominator":13,"numerator":525},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","peer_count":82,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"4.0","peer_count":49,"percentile":1.04,"percentile_exact":{"denominator":24,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":22.37,"percentile_exact":{"denominator":19,"numerator":425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":78,"percentile":32.47,"percentile_exact":{"denominator":77,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","peer_count":24,"percentile":30.43,"percentile_exact":{"denominator":23,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"2.0","peer_count":50,"percentile":26.53,"percentile_exact":{"denominator":49,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1.1","peer_count":27,"percentile":3.85,"percentile_exact":{"denominator":13,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665}],"release_date":"2026-07-30","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":29.23,"score_exact":{"denominator":1125107673161472,"numerator":32892305162137165},"source_id":"thinkingmachines/inkling-small"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":177,"numerator":6250},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2200},"family":"GPQA Diamond","row_ids":["7c995d56b5fa39eef927afa2cc879ab9"],"score":37.29},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":100},"family":"HLE","row_ids":["12e6c7bd6a0e1d4d1047be66dc2b360a"],"score":33.33}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":37.29,"percentile_exact":{"denominator":59,"numerator":2200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond","peer_count":64,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115}],"score":35.31,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":32683129},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":275},"family":"tau3-Banking","row_ids":["69ddb67fade6ab2b7cbae1dad6474f61"],"score":21.15},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2200},"family":"GDPval-AA","row_ids":["d1bad78363493513c436f20f24fa1380"],"score":27.16},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["27a6f0107d1b80cc1605307626b6cc35"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":400},"family":"AutomationBench","row_ids":["5cc05f32d32a7a675fc47d1b76ed156a"],"score":21.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1800},"family":"AA-Briefcase","row_ids":["1e5075fa69ac15167616fbc24b2c2ddf"],"score":23.38}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":21.15,"percentile_exact":{"denominator":13,"numerator":275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v2.1","peer_count":82,"percentile":27.16,"percentile_exact":{"denominator":81,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":21.05,"percentile_exact":{"denominator":19,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v1.1","peer_count":78,"percentile":23.38,"percentile_exact":{"denominator":77,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73}],"score":21.22,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1175},"family":"GDP.pdf","row_ids":["c1a1527152276d41667adbe5c22b811f"],"score":30.92}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"score":30.92,"weight":20}],"eligible":true,"id":"m-6ac02cf910375655","in_window":true,"name":"LongCat 2.0","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-29","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":86,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":37.29,"percentile_exact":{"denominator":59,"numerator":2200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond","peer_count":64,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":21.15,"percentile_exact":{"denominator":13,"numerator":275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v2.1","peer_count":82,"percentile":27.16,"percentile_exact":{"denominator":81,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":21.05,"percentile_exact":{"denominator":19,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v1.1","peer_count":78,"percentile":23.38,"percentile_exact":{"denominator":77,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"release_date":"2026-06-29","release_status":"published","release_verified":true,"result_updated_at":null,"score":29.02,"score_exact":{"denominator":1272485214,"numerator":36930684955},"source_id":"meituan/longcat-2.0"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":96088925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2900},"family":"GPQA Diamond","row_ids":["ff2b5baedd439ff80001794125500940"],"score":49.15},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["f04dc7e5afab13c2812e16f8a7f09157"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1700},"family":"HLE","row_ids":["67c349c9897445ce219d7f8acc0e4dbe"],"score":26.98},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":400},"family":"CritPt","row_ids":["142184a428fc22f5f377c4fcd96f6a10"],"score":23.53}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":49.15,"percentile_exact":{"denominator":59,"numerator":2900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","peer_count":64,"percentile":26.98,"percentile_exact":{"denominator":63,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","peer_count":52,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07}],"score":29.24,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":22547889,"numerator":644722264},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":750},"family":"tau3-Banking","row_ids":["4a499df1f84b1eaf0a172cfe97a95f4c"],"score":57.69},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2000},"family":"GDPval-AA","row_ids":["cab0e8633e6425575d15983329eb67d3"],"score":24.69},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["af8dab2781940ab31c2449f02ac9e502"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":500},"family":"AutomationBench","row_ids":["af88768655cfd944c0cc2a22a6f5661c"],"score":26.32},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":100},"family":"AA-Briefcase","row_ids":["6b81d5a0603a102ca44c8276f19acede"],"score":14.29},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":250},"family":"EnterpriseOps-Gym","row_ids":["6751eebce76aacc484338eb2b0237376"],"score":27.78},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":400},"family":"AA-AnalystAgent","row_ids":["809f69850cfb92e35fdef7e6fcd0b405"],"score":17.39}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","peer_count":82,"percentile":24.69,"percentile_exact":{"denominator":81,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":78,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","peer_count":24,"percentile":17.39,"percentile_exact":{"denominator":23,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2}],"score":28.59,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":900},"family":"FrontierCode","row_ids":["eae24b4f084dd40f578e44715aebd484"],"score":18.37}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","peer_count":50,"percentile":18.37,"percentile_exact":{"denominator":49,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818}],"score":18.37,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":137425},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1625},"family":"GDP.pdf","row_ids":["0849343c6ca0b1de586bda128e33678b"],"score":42.76},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1350},"family":"AA-Omniscience","row_ids":["ce91d9d68242cca613b5b932f9a344d4","934c0d476e0913dedfc11ab20974d89e"],"score":25.47}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548}],"score":34.12,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":450},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":450},"family":"MMMU-Pro","row_ids":["681af2b5c57d4ca46641c38951c3d479"],"score":26.47}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069}],"score":26.47,"weight":10}],"eligible":true,"id":"m-bf7f21a45c57e0f7","in_window":true,"name":"Muse Glimmer","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":87,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":49.15,"percentile_exact":{"denominator":59,"numerator":2900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","peer_count":64,"percentile":26.98,"percentile_exact":{"denominator":63,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","peer_count":52,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","peer_count":82,"percentile":24.69,"percentile_exact":{"denominator":81,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":78,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","peer_count":24,"percentile":17.39,"percentile_exact":{"denominator":23,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","peer_count":50,"percentile":18.37,"percentile_exact":{"denominator":49,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069}],"release_date":"2026-08-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":27.6,"score_exact":{"denominator":19177971701616,"numerator":529374170552879},"source_id":"meta/muse-glimmer-30b"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":63,"numerator":1000},"families":[{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1000},"family":"HLE","row_ids":["11ffdf294b4e40190b29f471a44eb232"],"score":15.87}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","peer_count":64,"percentile":15.87,"percentile_exact":{"denominator":63,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":15.87,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":248026779,"numerator":8390522060},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"APEX-Agents","row_ids":["ca1ee026c3875e1f8a6f67a549e74fe8"],"score":25.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1000},"family":"tau3-Banking","row_ids":["c993b26e42ea6045232f1d9e621ef2fa"],"score":76.92},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3500},"family":"GDPval-AA","row_ids":["e73dd8b4c1ab25719febf181d0bca142"],"score":43.21},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["d43f0f17f32dc60cbed0bdb79fe20eb1"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":525},"family":"AutomationBench","row_ids":["7fbcd0afe5186232a5dba432e082b6ef"],"score":27.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2900},"family":"AA-Briefcase","row_ids":["11eea946a9e871701b9adaec2c3d14b2"],"score":37.66},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":300},"family":"AA-AnalystAgent","row_ids":["c283db79a5fa78b491d2c36ac505fd73"],"score":13.04}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","peer_count":82,"percentile":43.21,"percentile_exact":{"denominator":81,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":27.63,"percentile_exact":{"denominator":19,"numerator":525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","peer_count":78,"percentile":37.66,"percentile_exact":{"denominator":77,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","peer_count":24,"percentile":13.04,"percentile_exact":{"denominator":23,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625}],"score":33.83,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1075},"family":"GDP.pdf","row_ids":["0e26cf9d2beafb47ed1ed89556040c14"],"score":28.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":28.29,"weight":20}],"eligible":true,"id":"m-95ce1c534519e66a","in_window":true,"name":"Ling-3.0-flash-Fin","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":88,"records":[{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","peer_count":64,"percentile":15.87,"percentile_exact":{"denominator":63,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","peer_count":82,"percentile":43.21,"percentile_exact":{"denominator":81,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":27.63,"percentile_exact":{"denominator":19,"numerator":525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","peer_count":78,"percentile":37.66,"percentile_exact":{"denominator":77,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","peer_count":24,"percentile":13.04,"percentile_exact":{"denominator":23,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"release_date":"2026-09-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":25.83,"score_exact":{"denominator":1736187453,"numerator":44851731725},"source_id":"inclusionai/ling-3.0-flash-fin"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":39117,"numerator":787925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1850},"family":"GPQA Diamond","row_ids":["885831acf33b9d6286fb559cc7b4ec0e"],"score":31.36},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d166f2e488fa71112bb60e275a7535f1"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["3441fc8b1a39080bc19e5b920fef2c87"],"score":11.76}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":20.14,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1100385,"numerator":49356484},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":800},"family":"tau3-Banking","row_ids":["33c50c79dd62f0151f2db3df92a057bc"],"score":61.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2900},"family":"GDPval-AA","row_ids":["4aa9f9d8e559b9430ab1f871a4ca4271"],"score":35.8},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["5a757d2c757812d80a306fe7007417cd"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":750},"family":"AutomationBench","row_ids":["b87ec7052a30aebc07b1b22c9deef3cd"],"score":39.47},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":500},"family":"AA-Briefcase","row_ids":["6d3a1ea10b80de3baf1922e2ef5fc22e"],"score":45.45}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","peer_count":82,"percentile":35.8,"percentile_exact":{"denominator":81,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","peer_count":78,"percentile":45.45,"percentile_exact":{"denominator":11,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51}],"score":44.85,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":800},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":800},"family":"FrontierCode","row_ids":["d64bab068e454f76d91e1fb18d73dd68"],"score":16.33}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","peer_count":50,"percentile":16.33,"percentile_exact":{"denominator":49,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525}],"score":16.33,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":625},"family":"GDP.pdf","row_ids":["48f1c6876fabfb820ee570e743c9611a"],"score":16.45}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"score":16.45,"weight":20}],"eligible":true,"id":"m-c58d5bdb4d2e38bc","in_window":true,"name":"K2 Horizon 7B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":89,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","peer_count":82,"percentile":35.8,"percentile_exact":{"denominator":81,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","peer_count":78,"percentile":45.45,"percentile_exact":{"denominator":11,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","peer_count":50,"percentile":16.33,"percentile_exact":{"denominator":49,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":25.34,"score_exact":{"denominator":194690237742,"numerator":4933003274473},"source_id":"published/aa:k2-horizon-7b"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":83020675},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2650},"family":"GPQA Diamond","row_ids":["c948234960090e8c9ef801d4c2e468fc"],"score":44.92},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["5d3cac863b9cb777dcf0429fca5619c7"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1150},"family":"HLE","row_ids":["1e3489d548cb804c3cfe4ff52df120e8"],"score":18.25},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":350},"family":"CritPt","row_ids":["55904790fd4c7a2e974218f474b3d773"],"score":20.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":44.92,"percentile_exact":{"denominator":59,"numerator":2650},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04}],"score":25.27,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":73359,"numerator":1386274},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":225},"family":"tau3-Banking","row_ids":["77cc539d6e52970112947681c17a7cf5"],"score":17.31},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":700},"family":"GDPval-AA","row_ids":["3eb5f8707daab0412d08dc253ada641e"],"score":25.93},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["017b50b4eb6f553388684608c12adbeb"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":375},"family":"AutomationBench","row_ids":["cf48341c74514d20a48b2f428df05755"],"score":19.74},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":200},"family":"AA-Briefcase","row_ids":["42d65fa891e8c3a4f35b4ea031798999"],"score":18.18}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","peer_count":82,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":78,"percentile":18.18,"percentile_exact":{"denominator":11,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09}],"score":18.9,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":500},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":500},"family":"FrontierCode","row_ids":["a86b3f52322966e22c9f2052d8fbc893"],"score":10.2}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","peer_count":50,"percentile":10.2,"percentile_exact":{"denominator":49,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404}],"score":10.2,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":119725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":825},"family":"GDP.pdf","row_ids":["7e1c74ae3bf26c2c5b56bd7d43f2a5df"],"score":21.71},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2000},"family":"AA-Omniscience","row_ids":["6b5784505206d57d0bba2ceb5b8de557","3bcf44016a4447d1427e1bf692c1e1bc"],"score":37.74}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963}],"score":29.72,"weight":20}],"eligible":true,"id":"m-c46d8f7f1c3144cb","in_window":true,"name":"K-EXAONE 2.0 0803","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":90,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":44.92,"percentile_exact":{"denominator":59,"numerator":2650},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","peer_count":82,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":78,"percentile":18.18,"percentile_exact":{"denominator":11,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","peer_count":50,"percentile":10.2,"percentile_exact":{"denominator":49,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":21.14,"score_exact":{"denominator":13758110133768,"numerator":290851131141355},"source_id":"published/aa:k-exaone-2-0-0803"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":1500},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1500},"family":"GPQA Diamond","row_ids":["3e2102f0b8440ab308d1e2fedeb7c1ef"],"score":25.42}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":25.42,"percentile_exact":{"denominator":59,"numerator":1500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434}],"score":25.42,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3081078,"numerator":55952933},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":1225},"family":"tau3-Banking","row_ids":["e330721d27c5f1f48579252fe1951f5e"],"score":47.12},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1300},"family":"GDPval-AA","row_ids":["f9e9baed830cb24bcff4431e5893ddf5"],"score":16.05},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["9fef6d5b9c29a36ef4ec13143b2d54ae"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":25},"family":"AutomationBench","row_ids":["a389be9490eced3c5e8431f3790a6ecd"],"score":1.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1000},"family":"AA-Briefcase","row_ids":["662e05367e55e07efb079daf8cc94493"],"score":12.99}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","peer_count":82,"percentile":16.05,"percentile_exact":{"denominator":81,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":78,"percentile":12.99,"percentile_exact":{"denominator":77,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91}],"score":18.16,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":75025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":125},"family":"GDP.pdf","row_ids":["7738f3afb94846ed2b13073b7d2cb3b1"],"score":3.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["97a4da3275fb19d0e062184a2de4a918","ff2df3ac7046d17ae8e64e925f03f746"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413}],"score":18.63,"weight":20}],"eligible":true,"id":"m-6b5176ed51518e35","in_window":true,"name":"Ling 3.0 Tiny","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":91,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":25.42,"percentile_exact":{"denominator":59,"numerator":1500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","peer_count":82,"percentile":16.05,"percentile_exact":{"denominator":81,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":78,"percentile":12.99,"percentile_exact":{"denominator":77,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413}],"release_date":"2026-08-06","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.89,"score_exact":{"denominator":134883432684,"numerator":2817358651405},"source_id":"published/aa:ling-3-0-tiny"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":49583375},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1700},"family":"GPQA Diamond","row_ids":["f6a22fbde90853e88aebc327f5973f62"],"score":28.81},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["1b904acff7083099b3a91e7b6c5b575a"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":650},"family":"HLE","row_ids":["5c0267c76de42807699083ce1e0581c4"],"score":10.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":200},"family":"CritPt","row_ids":["98267d8baf13117a13191a66537ee6eb"],"score":3.92}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":28.81,"percentile_exact":{"denominator":59,"numerator":1700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","peer_count":52,"percentile":3.92,"percentile_exact":{"denominator":51,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005}],"score":15.09,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1450449,"numerator":28030885},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":350},"family":"tau3-Banking","row_ids":["d0f6a7a8e824abf26de5a2c3eb1a8991"],"score":26.92},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":200},"family":"GDPval-AA","row_ids":["641dcee193d24a80d7b9a09327dfcf69"],"score":22.22},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["47e9d9d6a84d129959dab9eed70b71bc"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"AutomationBench","row_ids":["b05e70f346089460a993097793ed0c46"],"score":25.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1300},"family":"AA-Briefcase","row_ids":["671c80eaf4d486bab2d265a1c67045fd"],"score":16.88},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"EnterpriseOps-Gym","row_ids":["da0d77fdb30d1e419d687691f9d0af7b"],"score":22.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":200},"family":"AA-AnalystAgent","row_ids":["ef6364df2840f04f7c6f211bd995980d"],"score":8.7}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","peer_count":82,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":78,"percentile":16.88,"percentile_exact":{"denominator":77,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","peer_count":24,"percentile":8.7,"percentile_exact":{"denominator":23,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125}],"score":19.33,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":1000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1000},"family":"FrontierCode","row_ids":["c2dd8ca14021780f9edc8dc0975ff8ae"],"score":20.41}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","peer_count":50,"percentile":20.41,"percentile_exact":{"denominator":49,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939}],"score":20.41,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":106,"numerator":3825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"IFBench","row_ids":["5fbfa4015dcb4ab37c1080316055d055"],"score":75.0},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"GDP.pdf","row_ids":["2bd6dbb1c5200680f554dfce1a26feb0"],"score":12.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1100},"family":"AA-Omniscience","row_ids":["2b5b8d191f3e5e3bfa651cee2bbbd819","d7d18670af809e53c710911ecd277233"],"score":20.75}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378}],"score":36.08,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":150},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":150},"family":"MMMU-Pro","row_ids":["ce0faae345d00b20396f1024d947c26c"],"score":8.82}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798}],"score":8.82,"weight":10}],"eligible":true,"id":"m-5c18cc0e1d3391a9","in_window":true,"name":"Mistral Medium 3.5","official_evidence":{"artifact_sha256":"2f83ab54cf0df8335b080c25c95cc6852ada259bf85c12135f848a991e748f85","basis":"official_release_page","date":"2026-05-22","sha256":"2f83ab54cf0df8335b080c25c95cc6852ada259bf85c12135f848a991e748f85","url":"https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5/"},"primary_profile_id":"native","profile_id":"native","rank":92,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":28.81,"percentile_exact":{"denominator":59,"numerator":1700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","peer_count":52,"percentile":3.92,"percentile_exact":{"denominator":51,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","peer_count":82,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":78,"percentile":16.88,"percentile_exact":{"denominator":77,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","peer_count":24,"percentile":8.7,"percentile_exact":{"denominator":23,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","peer_count":50,"percentile":20.41,"percentile_exact":{"denominator":49,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798}],"release_date":"2026-05-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.78,"score_exact":{"denominator":411223564752,"numerator":8547240422815},"source_id":"mistral/mistral-medium-3-5"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":474012,"numerator":12975565},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2600},"family":"GDPval-AA","row_ids":["a6f95ea45eef0a446f5d6ed10bf3ca1b"],"score":32.1},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["614dafc454b6d64e61b2f25573ec4717"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":625},"family":"AutomationBench","row_ids":["d9c233b856dc8ecc623629216c3b1c31"],"score":32.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2400},"family":"AA-Briefcase","row_ids":["e9c29563b5cc25446296aa38a0ecd4ae"],"score":31.17}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","peer_count":82,"percentile":32.1,"percentile_exact":{"denominator":81,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","peer_count":78,"percentile":31.17,"percentile_exact":{"denominator":77,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94}],"score":27.37,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":2,"numerator":25},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"GDP.pdf","row_ids":["b6eecb655a01ff6addbb4005e50d5c51"],"score":12.5}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028}],"score":12.5,"weight":20}],"eligible":true,"id":"m-3ad11f2b6253b364","in_window":true,"name":"Nemotron 3.5 Bolt 74B-A9B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":93,"records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","peer_count":82,"percentile":32.1,"percentile_exact":{"denominator":81,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","peer_count":78,"percentile":31.17,"percentile_exact":{"denominator":77,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.76,"score_exact":{"denominator":4266108,"numerator":88578425},"source_id":"published/aa:nemotron-3-5-bolt-74b-a9b"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":1000},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1000},"family":"GPQA Diamond","row_ids":["9804205b8a1d620728568c184381a38c"],"score":16.95}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":16.95,"percentile_exact":{"denominator":59,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869}],"score":16.95,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":27,"numerator":200},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":200},"family":"GDPval-AA","row_ids":["4fa0d328e940e0ba1d47d7e70c6e0521"],"score":7.41}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","peer_count":82,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5}],"score":7.41,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":175},"family":"IFBench","row_ids":["8e2367487c04d7937c7f24678241f073"],"score":43.75}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252}],"score":43.75,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":200},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":200},"family":"MMMU-Pro","row_ids":["d47a048ab5b9e3796a7da71c7471a272"],"score":11.76}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144}],"score":11.76,"weight":10}],"eligible":true,"id":"m-fe0a26f770e1e011","in_window":true,"name":"DiffusionGemma 26B A4B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":94,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":16.95,"percentile_exact":{"denominator":59,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","peer_count":82,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144}],"release_date":"2026-06-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.02,"score_exact":{"denominator":433296,"numerator":8674375},"source_id":"published/aa:diffusiongemma-26b-a4b"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":19,"numerator":200},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":0},"family":"FrontierMath","row_ids":["040f48951e0ebb9ff6bfbde83c816f48"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":0},"family":"GPQA Diamond","row_ids":["29bbe2be4ccedeb019b3ccf6a3404eb9"],"score":0.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["540adb8175ab28d08c531a963e348646"],"score":31.58}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v2","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":10.53,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":5,"exact":{"denominator":3,"numerator":124},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"APEX-Agents","row_ids":["9584a53c71e055a6a275ffd8975b0b25"],"score":40.0},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":275},"family":"Terminal-Bench","row_ids":["2539da8c839981606943a4efe846b325"],"score":45.83},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"AutomationBench","row_ids":["6bd66300b839f4ff34d5fd8249350449"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"AA-Briefcase","row_ids":["52ae106aa8230eabcaac4637b6d547ea"],"score":50.0},{"benchmark_version":"1000 tasks","exact":{"denominator":3,"numerator":100},"family":"MCP Atlas","row_ids":["4cec2c81b7ca1a60e4ad36412d5896a7"],"score":33.33}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","peer_count":49,"percentile":45.83,"percentile_exact":{"denominator":6,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2}],"score":41.33,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["dee93fd91e74e8ba6728dba7ca4fd982"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":459,"numerator":11950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"IFBench","row_ids":["0a5ea9e2a9d5f26a6a034bb3853216a5"],"score":22.22},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"GDP.pdf","row_ids":["dc77f0ff3c077230aef1b2b8cb04ec92"],"score":0.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":950},"family":"SimpleQA Verified","row_ids":["dbd9d582709dbd41201aa5f60842dfb0"],"score":55.88}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53}],"score":26.03,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":9,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["2c5710356bbeeb960760bc793dbd6d80"],"score":11.11},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["24b2d438f8c9ef458157a109e7bcb701"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","peer_count":28,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9}],"score":5.56,"weight":10}],"eligible":true,"id":"m-cccbc1ce3042d216","in_window":true,"name":"Claude Opus 4.8","official_evidence":{"artifact_sha256":"f43b616507c976c560385970d34c456a9634c2e81776613c8670c64ab10e5e35","basis":"official_release_page","date":"2026-05-28","sha256":"f43b616507c976c560385970d34c456a9634c2e81776613c8670c64ab10e5e35","url":"https://www.anthropic.com/news/claude-opus-4-8"},"primary_profile_id":"native","profile_id":"native","rank":95,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v2","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","peer_count":49,"percentile":45.83,"percentile_exact":{"denominator":6,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","peer_count":28,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9}],"release_date":"2026-05-28","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":18.73,"score_exact":{"denominator":8721,"numerator":163322},"source_id":"anthropic/claude-opus-4-8"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":117351,"numerator":1674775},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1100},"family":"GPQA Diamond","row_ids":["c96afdd76392eb5d2c9b157b68b7de9e"],"score":18.64},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d7eec79b902bf9a304fdfc120e9aa124"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":350},"family":"CritPt","row_ids":["9fef9d533791c4fe2ca276032af2b566"],"score":6.86}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":18.64,"percentile_exact":{"denominator":59,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01}],"score":14.27,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":42819149},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":550},"family":"tau3-Banking","row_ids":["7b39b47195e57aecded4286410a368f3"],"score":42.31},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2300},"family":"GDPval-AA","row_ids":["eada0c68a6bf5686fb53166d5b815ba4"],"score":28.4},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["585ae90ea798ef43916073acc85d178e"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":575},"family":"AutomationBench","row_ids":["eb41669f3fb02b0d8fee80ec8c840979"],"score":30.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1900},"family":"AA-Briefcase","row_ids":["ed41f889fc4e7786b137bd6b76a3da51"],"score":24.68}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","peer_count":82,"percentile":28.4,"percentile_exact":{"denominator":81,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":30.26,"percentile_exact":{"denominator":19,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","peer_count":78,"percentile":24.68,"percentile_exact":{"denominator":77,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97}],"score":27.79,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":600},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":600},"family":"FrontierCode","row_ids":["908a5049e650c9f2841b87a654313ae7"],"score":12.24}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","peer_count":50,"percentile":12.24,"percentile_exact":{"denominator":49,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798}],"score":12.24,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":625},"family":"GDP.pdf","row_ids":["e8733c3c5f23f9a2dee5ae8306fe34a6"],"score":16.45}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"score":16.45,"weight":20}],"eligible":true,"id":"m-26bf3900b3ceb955","in_window":true,"name":"K2 Horizon 3.7B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":96,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":18.64,"percentile_exact":{"denominator":59,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","peer_count":82,"percentile":28.4,"percentile_exact":{"denominator":81,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":30.26,"percentile_exact":{"denominator":19,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","peer_count":78,"percentile":24.68,"percentile_exact":{"denominator":77,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","peer_count":50,"percentile":12.24,"percentile_exact":{"denominator":49,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":18.06,"score_exact":{"denominator":97345118871,"numerator":1758167195585},"source_id":"published/aa:k2-horizon-3-7b"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":41628025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":900},"family":"GPQA Diamond","row_ids":["651c19da954966c3a71036c8d3db174d"],"score":15.25},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["47e73ec519c6695efcb0b041ef5795ce"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":400},"family":"HLE","row_ids":["db2ce17572462c1eb76d15ea72c97e2b"],"score":6.35},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["d5d41ba5c0f73bb7b22fdca4916cb1b7"],"score":11.76}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":15.25,"percentile_exact":{"denominator":59,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","peer_count":64,"percentile":6.35,"percentile_exact":{"denominator":63,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":12.67,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":118503,"numerator":1813883},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":25},"family":"tau3-Banking","row_ids":["767893d84c1938b3d363a56f8204eb3b"],"score":25.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1400},"family":"GDPval-AA","row_ids":["a655895858c8427a30e33e8793a2117d"],"score":17.28},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["027860a90cc2579f5f5af8c36e50c87e"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":200},"family":"AutomationBench","row_ids":["354c0d77cbcd6264333d316fdf9f676b"],"score":10.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":800},"family":"AA-Briefcase","row_ids":["afdcb1efb0aca22bea9c16a3d016167d"],"score":10.39}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","peer_count":82,"percentile":17.28,"percentile_exact":{"denominator":81,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":10.53,"percentile_exact":{"denominator":19,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":78,"percentile":10.39,"percentile_exact":{"denominator":77,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75}],"score":15.31,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":107275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":375},"family":"GDP.pdf","row_ids":["6e95a209bd026922387753b507be18e1"],"score":9.87},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["d19df36b3cd3a0e840c1256cd1c744fe","82eab72497730b1d30475a9ca4c7819b"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948}],"score":26.63,"weight":20}],"eligible":true,"id":"m-9453d55ff407aaee","in_window":true,"name":"Granite 4.2 30B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":97,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":15.25,"percentile_exact":{"denominator":59,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","peer_count":64,"percentile":6.35,"percentile_exact":{"denominator":63,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","peer_count":82,"percentile":17.28,"percentile_exact":{"denominator":81,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":10.53,"percentile_exact":{"denominator":19,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":78,"percentile":10.39,"percentile_exact":{"denominator":77,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":17.6,"score_exact":{"denominator":4586036711256,"numerator":80716639816145},"source_id":"published/aa:granite-4-2-30b"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":625},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":0},"family":"FrontierMath","row_ids":["0c9a231c5ad369432b2d4e6968b6bc7c"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":25},"family":"GPQA Diamond","row_ids":["710e7d489e0ed28e0e46334dd5f83f89"],"score":12.5},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["85fbb43a0855c52959cd8ec4413297f0"],"score":3.95}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v2","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"score":5.48,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":54,"numerator":1475},"families":[{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":25},"family":"Terminal-Bench","row_ids":["5bec3b19db8e7a3362bcd865766b2cc8"],"score":25.0},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"AutomationBench","row_ids":["51dfce6bb3b25136b0ea3545031f03b8"],"score":12.5},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":400},"family":"AA-Briefcase","row_ids":["06516697c2ae14f64607af00c620c8a8"],"score":44.44}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","peer_count":49,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798}],"score":27.31,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":25},"family":"DeepSWE","row_ids":["b632e7c137b19cc34931e9481ad27bf9"],"score":12.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2}],"score":12.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"GDP.pdf","row_ids":["706d4a87bdc7adf93c73d7e4099021c8"],"score":25.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7}],"score":25.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["6698824b06580d547cc03c11c24e9e05"],"score":3.7},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["b27c843f9cde1b388c12022729f0ceb0"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","peer_count":28,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"reference_revision":"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4}],"score":1.85,"weight":10}],"eligible":true,"id":"m-65a8fc1dec61133f","in_window":true,"name":"GPT-5.6 Luna","official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profile_id":"native","rank":98,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v2","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","peer_count":49,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","peer_count":28,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"reference_revision":"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":"2026-07-09","score":15.88,"score_exact":{"denominator":684,"numerator":10865},"source_id":"openai/gpt-5.6-luna"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":52678025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1600},"family":"GPQA Diamond","row_ids":["02699c8c436679b49b6319aa832d4d14"],"score":27.12},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["ac28090aae70acdb086bd22103c739d0"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":500},"family":"HLE","row_ids":["e12a9fbbd9fdb2bbd3147c3b6679efcd"],"score":7.94},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["2f96f5af894d02c94d57177465def23b"],"score":11.76}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":27.12,"percentile_exact":{"denominator":59,"numerator":1600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond","peer_count":64,"percentile":7.94,"percentile_exact":{"denominator":63,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":16.03,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":22949483},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":200},"family":"tau3-Banking","row_ids":["612f78abda8cf62d29f104d80c127636"],"score":15.38},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":500},"family":"GDPval-AA","row_ids":["1694dd426967f516c30dcdcc5616df9b"],"score":18.52},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["803b4920eccad03514634ebd3575a2d8"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":150},"family":"AutomationBench","row_ids":["ec3633dbe4490f5d8334e3da32d471af"],"score":7.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1200},"family":"AA-Briefcase","row_ids":["47de14e8ef5807902d31f4e62879bab3"],"score":15.58},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"EnterpriseOps-Gym","row_ids":["80ae66f2b7393fe1f51bab7846dd0a10"],"score":0.0}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2.1","peer_count":82,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":78,"percentile":15.58,"percentile_exact":{"denominator":77,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA run","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537}],"score":14.9,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["c503c45f721e54502bd0c3c277c347cb"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":46875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":275},"family":"GDP.pdf","row_ids":["40a272cc2b1879bb5b0104ebb8833d18"],"score":14.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1700},"family":"AA-Omniscience","row_ids":["8698ac189bdac0fe4a6655a8ed1ade73","2184077575074c69e38c71abae562f14"],"score":32.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636}],"score":23.27,"weight":20}],"eligible":true,"id":"m-9ff6170077842b96","in_window":true,"name":"Nemotron 3.5 Lightning","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":99,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":27.12,"percentile_exact":{"denominator":59,"numerator":1600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond","peer_count":64,"percentile":7.94,"percentile_exact":{"denominator":63,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2.1","peer_count":82,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":78,"percentile":15.58,"percentile_exact":{"denominator":77,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA run","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636}],"release_date":"2026-08-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":14.44,"score_exact":{"denominator":41274330401304,"numerator":596156995131155},"source_id":"nvidia/nemotron-3.5-lightning"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1095276,"numerator":17204225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2000},"family":"GPQA Diamond","row_ids":["eebaf08b0e483d418dfe3796053cb2a9"],"score":33.9},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["b878256e74c2c14b269834f44b853cea"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":100},"family":"HLE","row_ids":["19f9062134531b120e5a3beb1b827183"],"score":4.76},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":350},"family":"CritPt","row_ids":["1ce53974b2fd41332a137498132cd83b"],"score":6.86}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":33.9,"percentile_exact":{"denominator":59,"numerator":2000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","peer_count":64,"percentile":4.76,"percentile_exact":{"denominator":21,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01}],"score":15.71,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":93810898},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":125},"family":"tau3-Banking","row_ids":["a7dc3e52deea9234fc190c75c4007564"],"score":9.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":500},"family":"GDPval-AA","row_ids":["72fabae9540ce0210203fd32c61249bb"],"score":6.17},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["c32f27bfb7b7fcf6e0fb4b8f49b0314a"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":175},"family":"AutomationBench","row_ids":["427f2b62a677b7b14b79a216cd5cce2a"],"score":9.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":300},"family":"AA-Briefcase","row_ids":["f3f72292d6ff516b21ceaec7da17a866"],"score":3.9}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":9.62,"percentile_exact":{"denominator":13,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","peer_count":82,"percentile":6.17,"percentile_exact":{"denominator":81,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":9.21,"percentile_exact":{"denominator":19,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":78,"percentile":3.9,"percentile_exact":{"denominator":77,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45}],"score":12.18,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":400},"family":"FrontierCode","row_ids":["e2d037eb56a1707481cab5a658eb199f"],"score":8.16}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"2.0","peer_count":50,"percentile":8.16,"percentile_exact":{"denominator":49,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101}],"score":8.16,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3021,"numerator":62900},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"IFBench","row_ids":["65c9161268988be090ca4411568e9e7b"],"score":37.5},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":375},"family":"GDP.pdf","row_ids":["96fc6325ca8f16276f80447a5bb8a745"],"score":9.87},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":800},"family":"AA-Omniscience","row_ids":["2d5bfdd87d275a4c6177037eb5a3ff85","30b64e451b40a0bb848f166f0f44b6dd"],"score":15.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037}],"score":20.82,"weight":20}],"eligible":true,"id":"m-2c518178af26d1d4","in_window":true,"name":"North Mini Code","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-09","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":100,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":33.9,"percentile_exact":{"denominator":59,"numerator":2000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","peer_count":64,"percentile":4.76,"percentile_exact":{"denominator":21,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":9.62,"percentile_exact":{"denominator":13,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","peer_count":82,"percentile":6.17,"percentile_exact":{"denominator":81,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":9.21,"percentile_exact":{"denominator":19,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":78,"percentile":3.9,"percentile_exact":{"denominator":77,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"2.0","peer_count":50,"percentile":8.16,"percentile_exact":{"denominator":49,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037}],"release_date":"2026-06-09","release_status":"published","release_verified":true,"result_updated_at":null,"score":14.19,"score_exact":{"denominator":3174948492408,"numerator":45043683141617},"source_id":"cohere/north-mini-code-1-0"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":67424575},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2100},"family":"GPQA Diamond","row_ids":["13075042d3067db5be44e3af38eac3de"],"score":35.59},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["705b0d6767e14eda4ee6da31459437c6"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":850},"family":"HLE","row_ids":["48d08d1a4f13b1f74f0a43f0e70c78cb"],"score":13.49},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":800},"family":"CritPt","row_ids":["bfeb15a230a3fe93a4b8668ae5c71e33"],"score":15.69}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":35.59,"percentile_exact":{"denominator":59,"numerator":2100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","peer_count":52,"percentile":15.69,"percentile_exact":{"denominator":51,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025}],"score":20.52,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2567565,"numerator":38748266},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":100},"family":"tau3-Banking","row_ids":["d4ac6d97f1d2caee1aac75fe86cb1d41"],"score":7.69},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":400},"family":"GDPval-AA","row_ids":["a7f8996adb7838234a9dd47a6d848a6c"],"score":14.81},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["ed0927341bbf29454c5f15ec4d2315cd"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":250},"family":"AutomationBench","row_ids":["c9594ae7e63800079a6fe5e87536af10"],"score":13.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":600},"family":"AA-Briefcase","row_ids":["d623fd47b28047c5307a67919c5c6fae"],"score":7.79}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","peer_count":82,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":13.16,"percentile_exact":{"denominator":19,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":78,"percentile":7.79,"percentile_exact":{"denominator":77,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2}],"score":15.09,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["c8987663cd611e1b09bc6211e110c938"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":81325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":25},"family":"IFBench","row_ids":["3f2a6e6ace968612fc72d2719d04f5ee"],"score":6.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":100},"family":"GDP.pdf","row_ids":["6612c73a65ff838439bc02c707904f4b"],"score":5.26},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2600},"family":"AA-Omniscience","row_ids":["0ca7e9728f633cc599fad0ed12a6097b","862933c1b9c8be1d0cbb4bb130b222e9"],"score":49.06}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","peer_count":17,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625}],"score":20.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":50},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":50},"family":"MMMU-Pro","row_ids":["b79c49748f47d291a424306d40dc6469"],"score":2.94}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532}],"score":2.94,"weight":10}],"eligible":true,"id":"m-8877a9675c79fcad","in_window":true,"name":"Command A+","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-20","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":101,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":35.59,"percentile_exact":{"denominator":59,"numerator":2100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","peer_count":52,"percentile":15.69,"percentile_exact":{"denominator":51,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","peer_count":82,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":13.16,"percentile_exact":{"denominator":19,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":78,"percentile":7.79,"percentile_exact":{"denominator":77,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","peer_count":17,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532}],"release_date":"2026-05-20","release_status":"published","release_verified":true,"result_updated_at":null,"score":13.85,"score_exact":{"denominator":15286789037520,"numerator":211678630100407},"source_id":"cohere/command-a-plus-05-2026"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3,"numerator":50},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":50},"family":"GPQA Diamond","row_ids":["55c9779a030f53f08065b75a89097776"],"score":16.67}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2}],"score":16.67,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3,"numerator":55},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"APEX-Agents","row_ids":["6f3782a0a7f13ec12e5eae491c79c282"],"score":20.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["96fb24279652563e7d2bc82b71d0fc52"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":100},"family":"AA-Briefcase","row_ids":["5e2ad6a27f43522ccb21c4f280e4613e"],"score":33.33},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":20},"family":"MCP Atlas","row_ids":["d57170302e825b662bfe4acca3fbd398"],"score":20.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6}],"score":18.33,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["546ef51508bcfe588737e01c240c5015"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":17,"numerator":250},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":250},"family":"SimpleQA Verified","row_ids":["bcb4fb304c488f86e521408952f81841"],"score":14.71}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"score":14.71,"weight":20}],"eligible":true,"id":"m-5cf6d704eda6778e","in_window":true,"name":"GLM-5.2","official_evidence":{"artifact_sha256":"c6c90dc0946031acbba79d00549ddb918494022b454142617bed1e9492fdd97e","basis":"official_release_page","date":"2026-06-16","sha256":"c6c90dc0946031acbba79d00549ddb918494022b454142617bed1e9492fdd97e","url":"https://z.ai/blog/glm-5.2"},"primary_profile_id":"native","profile_id":"native","rank":102,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"release_date":"2026-06-16","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":12.99,"score_exact":{"denominator":102,"numerator":1325},"source_id":"zhipuai/glm-5.2"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":2394,"numerator":20875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["80525908530cba67da8858ed258e1742"],"score":3.95},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":850},"family":"HLE","row_ids":["c6b63609fd0ed9a9d1cc3300620f0885"],"score":13.49}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051}],"score":8.72,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1881,"numerator":23195},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":100},"family":"GDPval-AA","row_ids":["21dd109a3c14f1e9b389fa0a30a5ee0f"],"score":11.11},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["179a39191c3950939aafcb199439b020"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":300},"family":"AutomationBench","row_ids":["43181abb2675b0ef85c7558935e2c229"],"score":15.79},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":100},"family":"AA-Briefcase","row_ids":["c4bef304832e9ad8f150cf2ae6ad2c77"],"score":9.09}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","peer_count":82,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":15.79,"percentile_exact":{"denominator":19,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","peer_count":78,"percentile":9.09,"percentile_exact":{"denominator":11,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49}],"score":12.33,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":350},"family":"GDP.pdf","row_ids":["8f8d77f644fefbdcb7d35712e527c726"],"score":18.42}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036}],"score":18.42,"weight":20}],"eligible":true,"id":"m-b9d9f5989f59e750","in_window":true,"name":"Mercury 2.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-08","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":103,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","peer_count":82,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":15.79,"percentile_exact":{"denominator":19,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","peer_count":78,"percentile":9.09,"percentile_exact":{"denominator":11,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036}],"release_date":"2026-09-08","release_status":"published","release_verified":true,"result_updated_at":null,"score":12.78,"score_exact":{"denominator":40964,"numerator":523575},"source_id":"inception/mercury-2.5"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":531,"numerator":13000},"families":[{"benchmark_version":"2026","exact":{"denominator":3,"numerator":100},"family":"AIME","row_ids":["072b96c67ae19a20ee48a971117995c0"],"score":33.33},{"benchmark_version":"500-item subset","exact":{"denominator":3,"numerator":100},"family":"MATH-500","row_ids":["158a7a0bcbb4fdee990dff1e880fb156"],"score":33.33},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":400},"family":"GPQA Diamond","row_ids":["249b24c9d6325f053552fb911f7f4a84"],"score":6.78}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","included":true,"model_id":"m-161fc8601724c921","original_version_label":"2026","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":6.78,"percentile_exact":{"denominator":59,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313}],"score":24.48,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"4","exact":{"denominator":1,"numerator":0},"family":"BFCL","row_ids":["602ba59a6dfe14df55137c17751e7666"],"score":0.0}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","included":true,"model_id":"m-161fc8601724c921","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5}],"score":0.0,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"release_v6","exact":{"denominator":1,"numerator":0},"family":"LiveCodeBench","row_ids":["445a31daff84293600dd4a2900cbf97d"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":48,"numerator":775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":125},"family":"IFBench","row_ids":["950f9e2f9181b2c99ee5f8151838596d"],"score":31.25},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"Multi-IF","row_ids":["5b52e92a8499178ddfa84c801132f2f4"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":100},"family":"IFEval","row_ids":["0c588d45b55a074bc2915de39e2cf5ff"],"score":33.33},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"MMLU-Pro","row_ids":["0129fdb380d08e0198547b85719a77cd"],"score":0.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1}],"score":16.15,"weight":20}],"eligible":true,"id":"m-161fc8601724c921","in_window":true,"name":"LFM2.5-8B-A1B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":104,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","included":true,"model_id":"m-161fc8601724c921","original_version_label":"2026","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":6.78,"percentile_exact":{"denominator":59,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","included":true,"model_id":"m-161fc8601724c921","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1}],"release_date":"2026-05-28","release_status":"published","release_verified":true,"result_updated_at":null,"score":10.39,"score_exact":{"denominator":38232,"numerator":397175},"source_id":"published/aa:lfm2-5-8b-a1b"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":28072925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":750},"family":"GPQA Diamond","row_ids":["80f85d325d6536dbd1603ecf4678090b"],"score":12.71},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["10a24da4314ddad1e5487011d9ce22be"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":200},"family":"HLE","row_ids":["1c035cc3ccd2789e7031fcc995e521f7"],"score":3.17},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":50},"family":"CritPt","row_ids":["8e1dd162a62fd85fef38ef1d97e4e9bc"],"score":0.98}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","peer_count":64,"percentile":3.17,"percentile_exact":{"denominator":63,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":8.54,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":18519664},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":175},"family":"tau3-Banking","row_ids":["a1e714634bbaa4b289e4726c4bb2c9e9"],"score":13.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1000},"family":"GDPval-AA","row_ids":["5be65e1004d9e6f482549aefe65a50a4"],"score":12.35},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["f435dfe97ff078f0f04715c2a33b7561"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":275},"family":"AutomationBench","row_ids":["e3a197e716166eb9edc4c46f49e52295"],"score":14.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":500},"family":"AA-Briefcase","row_ids":["2e097f3022508687e5c06477d01a5253"],"score":6.49}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":13.46,"percentile_exact":{"denominator":13,"numerator":175},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","peer_count":82,"percentile":12.35,"percentile_exact":{"denominator":81,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":78,"percentile":6.49,"percentile_exact":{"denominator":77,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41}],"score":12.02,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["6a0e7ddb2b9af667874b506e526b6d54"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":35525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":25},"family":"GDP.pdf","row_ids":["8eaabce7efa57428ec82adf89ba50dbc"],"score":1.32},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["4a7b9f5169709891b5917ee76e2a8e1a","5f020104c70b0364755cf0f561bff098"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647}],"score":17.64,"weight":20}],"eligible":true,"id":"m-1d98b8595cf1927d","in_window":true,"name":"Granite 4.2 8B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":105,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","peer_count":64,"percentile":3.17,"percentile_exact":{"denominator":63,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":13.46,"percentile_exact":{"denominator":13,"numerator":175},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","peer_count":82,"percentile":12.35,"percentile_exact":{"denominator":81,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":78,"percentile":6.49,"percentile_exact":{"denominator":77,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":10.31,"score_exact":{"denominator":41274330401304,"numerator":425645973328915},"source_id":"ibm/granite-4.2-8b"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":57550},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":750},"family":"GPQA Diamond","row_ids":["63aced451500a12a72f3e068bcd6ff1f"],"score":12.71},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1150},"family":"HLE","row_ids":["04be87f5b8e53109685297a0e0936bc5"],"score":18.25}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057}],"score":15.48,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":118503,"numerator":588163},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":0},"family":"tau3-Banking","row_ids":["09489dd17969a18613dce0aff6c534ba"],"score":0.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":400},"family":"GDPval-AA","row_ids":["0753db930b01dc3794424662db360ca1"],"score":4.94},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["732ce0d4c4b93e00640afbdf97e81cd3"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":75},"family":"AutomationBench","row_ids":["b4943be7200f5ac57de9f15aacbca48d"],"score":3.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":200},"family":"AA-Briefcase","row_ids":["dda521b5d7d0c5ce2d8db12c4ea78f4c"],"score":2.6}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","peer_count":82,"percentile":4.94,"percentile_exact":{"denominator":81,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","peer_count":78,"percentile":2.6,"percentile_exact":{"denominator":77,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94}],"score":4.96,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":125},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":125},"family":"GDP.pdf","row_ids":["d70ece05a7cd9206504e0831b954b698"],"score":6.58}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012}],"score":6.58,"weight":20}],"eligible":true,"id":"m-7990b74b056a0b9a","in_window":true,"name":"Celeris-1","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-24","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":106,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","peer_count":82,"percentile":4.94,"percentile_exact":{"denominator":81,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","peer_count":78,"percentile":2.6,"percentile_exact":{"denominator":77,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012}],"release_date":"2026-07-24","release_status":"published","release_verified":true,"result_updated_at":null,"score":9.18,"score_exact":{"denominator":13983354,"numerator":128393905},"source_id":"published/aa:celeris-1"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":100},"family":"GPQA Diamond","row_ids":["990a7e694606915f66b15d0a96c06abb"],"score":1.69}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","included":true,"model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":1.69,"percentile_exact":{"denominator":59,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293}],"score":1.69,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":39,"numerator":119},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":25},"family":"tau3-Banking","row_ids":["57be0e2f641053db7912c66d73e8ba7b"],"score":1.92},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"GDPval-AA","row_ids":["710c23bbfe14fa2ea7bd1f3a24100a69"],"score":0.0},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["7ed61dbea4d39e1cec33826bfcfe4762"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["e01a487a4aa0d6109e212abc8d8c200e"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"AA-Briefcase","row_ids":["3f053e6933f1e5ef77b0bf87fdfb5476"],"score":0.0}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","included":true,"model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":1.92,"percentile_exact":{"denominator":13,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v2.1","peer_count":82,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","included":true,"model_id":"m-928c410ba1082113","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","included":true,"model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v1.1","peer_count":78,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0}],"score":3.05,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":125},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":125},"family":"GDP.pdf","row_ids":["b9fa019c4e1362fc93bc4088a9dc8dbc"],"score":3.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","included":true,"model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008}],"score":3.29,"weight":20}],"eligible":true,"id":"m-928c410ba1082113","in_window":true,"name":"K2 Horizon 0.9B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":107,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","included":true,"model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":1.69,"percentile_exact":{"denominator":59,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","included":true,"model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":1.92,"percentile_exact":{"denominator":13,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v2.1","peer_count":82,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","included":true,"model_id":"m-928c410ba1082113","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","included":true,"model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v1.1","peer_count":78,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","included":true,"model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":2.63,"score_exact":{"denominator":612066,"numerator":1612745},"source_id":"published/aa:k2-horizon-0-9b"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":200},"family":"GPQA Diamond","row_ids":["c0db3282b4d173e68108771ffab82d19"],"score":3.39}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":3.39,"percentile_exact":{"denominator":59,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505}],"score":3.39,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":81,"numerator":100},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":100},"family":"GDPval-AA","row_ids":["32822274d56779c973cdcf3ccb12841f"],"score":1.23}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","peer_count":82,"percentile":1.23,"percentile_exact":{"denominator":81,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32}],"score":1.23,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"IFBench","row_ids":["e359de5ff096ac45c506a2baeb714ed7"],"score":0.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","peer_count":17,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551}],"score":0.0,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["7a567c9ea42d069b6cee3f34d322f5a9"],"score":0.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087}],"score":0.0,"weight":10}],"eligible":true,"id":"m-0cd6b9e2d04f2c30","in_window":true,"name":"MiniCPM-V 4.6 1.3B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":108,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":3.39,"percentile_exact":{"denominator":59,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","peer_count":82,"percentile":1.23,"percentile_exact":{"denominator":81,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","peer_count":17,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087}],"release_date":"2026-05-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":1.45,"score_exact":{"denominator":19116,"numerator":27625},"source_id":"published/aa:minicpm-v-4-6-1-3b"}],"ranked_configuration_count":108,"ranked_model_count":108,"recent_model_count":108,"reference_populations":{"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35":{"benchmark":"HLE","benchmark_version":"Diamond","members":[{"configuration":"gpt-5.6-sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"156/5","observation_id":"52db35ab0a7c574170e312e2a679c471","value":31.2},{"configuration":"gpt-6.1-terra","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"357/10","observation_id":"9c15f08ff3f1d95aac1ab4460a6a090b","value":35.7},{"configuration":"gpt-6-sol","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"169/5","observation_id":"3ff9ef5159d8b127a6433bdd9ec141a5","value":33.8},{"configuration":"gemini-3.8-flash","model_id":"m-422e5ba87859dd55","normalized_unit_value":"343/10","observation_id":"08dc502d8d4aeb42de93130ed556d391","value":34.3},{"configuration":"gpt-6-astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"303/5","observation_id":"955cc9868301478204da3b4815a3ad3c","value":60.6},{"configuration":"claude-opus-5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"193/5","observation_id":"852fa04b612ed916ac880c4c5a9b046b","value":38.6},{"configuration":"muse-wonder-1.0","model_id":"m-68480fc0423c5034","normalized_unit_value":"373/10","observation_id":"c73228fc930f557212149ca3c921d3fb","value":37.3},{"configuration":"grok-4.7","model_id":"m-7187cf2b01d41935","normalized_unit_value":"117/5","observation_id":"da019b1bad73b978b878b2436172748d","value":23.4},{"configuration":"gemini-4-argon","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"266/5","observation_id":"c2167057895c8758046c1d7e9165f23f","value":53.2},{"configuration":"claude-fable-5-1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"513/10","observation_id":"16b64e0ecfeb64ebd3bc40ba58f5f691","value":51.3},{"configuration":"kimi-k3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"111/5","observation_id":"0e05cbe602caeb64b25d2230f75c2113","value":22.2},{"configuration":"inkling","model_id":"m-b439c804161fac26","normalized_unit_value":"56/5","observation_id":"6e97ec16100fc8b44c1121b2536cf94d","value":11.2},{"configuration":"muse-spark-1.3","model_id":"m-c0580b01221ed260","normalized_unit_value":"127/5","observation_id":"5a5d1e52735d8efc5b98d11b601491a8","value":25.4},{"configuration":"glm-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"82/5","observation_id":"f0c4c8c8946a2cc602d15226e2c11a49","value":16.4},{"configuration":"claude-opus-5-5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"55","observation_id":"eff89ef5cfa18da15eddc05443bb6b8d","value":55},{"configuration":"claude-sonnet-5-5","model_id":"m-d123610543e5f5ed","normalized_unit_value":"75/2","observation_id":"36082ffe886fbeb2647876856cf1e5c0","value":37.5},{"configuration":"muse-spark-open","model_id":"m-db208dec31031420","normalized_unit_value":"28","observation_id":"04518aae51e855cbdd25679b26d57f51","value":28}],"protocol_id":"de345739b732415f6fe3a72b","unit":"%"},"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5":{"benchmark":"GDPval-AA","benchmark_version":"2.1","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"163083/100","observation_id":"ac3f66c4919122d1f31985187718ed07","value":1630.83},{"configuration":"MiniCPM-V 4.6 1.3B","model_id":"m-0cd6b9e2d04f2c30","normalized_unit_value":"-9058/25","observation_id":"32822274d56779c973cdcf3ccb12841f","value":-362.32},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"160889/100","observation_id":"eaa5c4189ba4b51426be70de02910e04","value":1608.89},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"5231/25","observation_id":"a9600d8248b3ea039e7316ffc060f044","value":209.24},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"24937/50","observation_id":"5be65e1004d9e6f482549aefe65a50a4","value":498.74},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"147563/100","observation_id":"b0fdd931f8b81700b478f4129dec593b","value":1475.63},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"185709/100","observation_id":"fa97f8f99150ab638ce3d789e781973c","value":1857.09},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"7529/10","observation_id":"2a88f9d5ec82d3f04004cf3da53914a5","value":752.9},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"44441/50","observation_id":"eada0c68a6bf5686fb53166d5b815ba4","value":888.82},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"15931/50","observation_id":"72fabae9540ce0210203fd32c61249bb","value":318.62},{"configuration":"JT-4.1 Flash 236B A21B (Reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"133519/100","observation_id":"87c313476c9a420ad27b41db7d0250bf","value":1335.19},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"28314/25","observation_id":"fd7e6b8b2243df420c1327f60f09d326","value":1132.56},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"116241/100","observation_id":"3c9b5c57e54a04df82602451589cb092","value":1162.41},{"configuration":"Nemotron 3.5 Bolt 74B-A9B","model_id":"m-3ad11f2b6253b364","normalized_unit_value":"97903/100","observation_id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","value":979.03},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"51627/100","observation_id":"0bf079be27f6b7285cbdd993ff5b5a93","value":516.27},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"14299/20","observation_id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","value":714.95},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"144543/100","observation_id":"5e4d6ed857d58561b430218e5d2e3d99","value":1445.43},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"158369/100","observation_id":"b076d23f5b3bf4c32b53727e95233455","value":1583.69},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"76887/100","observation_id":"641dcee193d24a80d7b9a09327dfcf69","value":768.87},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"104601/100","observation_id":"ae60755488af790d34ab9ed3aa726374","value":1046.01},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"40339/25","observation_id":"42544e43967725b8e648b9b8615120d6","value":1613.56},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"155401/100","observation_id":"6a2341b888c3c4634de9d8d25fc69ba9","value":1554.01},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"37161/20","observation_id":"948bda542167bd107b882e3640f83a0c","value":1858.05},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"88517/50","observation_id":"5449b5fbd439528611d565a71ad28a45","value":1770.34},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"21916/25","observation_id":"d1bad78363493513c436f20f24fa1380","value":876.64},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"14204/25","observation_id":"f9e9baed830cb24bcff4431e5893ddf5","value":568.16},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"11897/10","observation_id":"95847d736743738d4e269c199a41aca9","value":1189.7},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"85921/50","observation_id":"4d4ab3fcd8c470a4ca605a1e507ade3e","value":1718.42},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"3874/5","observation_id":"c77b3a8d70f27e7507bc297339631f46","value":774.8},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"58471/50","observation_id":"2d97f2a57324a678b4a67b9a04a099de","value":1169.42},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"110573/100","observation_id":"2235f722890b82a4475fc455a58c66ab","value":1105.73},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"15473/50","observation_id":"0753db930b01dc3794424662db360ca1","value":309.46},{"configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","model_id":"m-7b2a8fe101678ebd","normalized_unit_value":"11214/25","observation_id":"fc44351a5b9d8e210a763f56506c0719","value":448.56},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"81533/50","observation_id":"feff6b1ffb0f6cb8ff2773defec53d29","value":1630.66},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"16739/10","observation_id":"8fccbdc41b807b2ebbe41cff024b0690","value":1673.9},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"10437/20","observation_id":"a7f8996adb7838234a9dd47a6d848a6c","value":521.85},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"36634/25","observation_id":"cfa7d9348c146c713f91a2534eae2b29","value":1465.36},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"-20109/50","observation_id":"710c23bbfe14fa2ea7bd1f3a24100a69","value":-402.18},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"58287/100","observation_id":"a655895858c8427a30e33e8793a2117d","value":582.87},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"27699/25","observation_id":"e73dd8b4c1ab25719febf181d0bca142","value":1107.96},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"176329/100","observation_id":"1bc96450387b9110981d343f4dfd20eb","value":1763.29},{"configuration":"Motif 3","model_id":"m-9e1e0f605d47cf02","normalized_unit_value":"56367/50","observation_id":"d1a5193baa7d56820cb5874aa60b7c4a","value":1127.34},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"16142/25","observation_id":"1694dd426967f516c30dcdcc5616df9b","value":645.68},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"82647/50","observation_id":"f04f5a2ca29d7ef8a19efd2486d1e259","value":1652.94},{"configuration":"Gemma 4 12B (Reasoning)","model_id":"m-a4be17f49ae915ce","normalized_unit_value":"43479/100","observation_id":"60ed1cde923cee15c4935672b177aa62","value":434.79},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"119667/100","observation_id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","value":1196.67},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"102069/100","observation_id":"d9fac2d0e8fa2fae504adca322b42dfa","value":1020.69},{"configuration":"G9v3-39A5B","model_id":"m-af0edba2f76c80d6","normalized_unit_value":"110937/100","observation_id":"5d726bb21290d73d92a00bf9993c5a4a","value":1109.37},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"54011/50","observation_id":"ff011af995e95f243ad573421cc4ba05","value":1080.22},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"29427/25","observation_id":"522de4d527da35b85f7ead36d5bab4a3","value":1177.08},{"configuration":"Solar Open2 250B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"978","observation_id":"7bf77397fcea6af7a97a2524f4c194bb","value":978},{"configuration":"Mercury 2.5","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"48179/100","observation_id":"21dd109a3c14f1e9b389fa0a30a5ee0f","value":481.79},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"79229/100","observation_id":"cab0e8633e6425575d15983329eb67d3","value":792.29},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"27737/25","observation_id":"9050cb18d0244c05aa4a3bdecef069eb","value":1109.48},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"4079/5","observation_id":"3eb5f8707daab0412d08dc253ada641e","value":815.8},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"102431/100","observation_id":"4aa9f9d8e559b9430ab1f871a4ca4271","value":1024.31},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"41422/25","observation_id":"8faa2a90e5a7df630665e90e7dc8da8c","value":1656.88},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"142673/100","observation_id":"022f63e78e1f4658e8af4b434e595fa3","value":1426.73},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"174577/100","observation_id":"0a034403545d9a6f875f18bdfdb8614b","value":1745.77},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"164937/100","observation_id":"3091d538e8d1e5edf6eea60314cea6a6","value":1649.37},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"186627/100","observation_id":"b44123b055ba24f02daafca81efcb87b","value":1866.27},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"119561/100","observation_id":"5a309e2224372571fd72b2e3de2fe6e1","value":1195.61},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"-53/100","observation_id":"7d56dec4d055d3d9b1237e11f0bf7ef4","value":-0.53},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"91891/50","observation_id":"c0ede537ba74a003e6de650e7903fd9e","value":1837.82},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"26766/25","observation_id":"3e2d18ecd2bf6e50f2e15788da94966d","value":1070.64},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"27301/20","observation_id":"3637cd472b3e658153b3601c1c953f3e","value":1365.05},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"48423/50","observation_id":"518222bbae6a6938c2fc68e86213e044","value":968.46},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"24999/20","observation_id":"7ee504c210f7e12209fb9a5011615203","value":1249.95},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"8447/5","observation_id":"82bb5fcac208ae67d3e4a7d3de124d5d","value":1689.4},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"35028/25","observation_id":"c140b9f2f3b8b047f022d22a6670f4c4","value":1401.12},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"43317/25","observation_id":"5c7968ea3f6404bb3d978285ac42aa05","value":1732.68},{"configuration":"Agnes 2.5 Pro Beta","model_id":"m-e060cf24e4176269","normalized_unit_value":"133533/100","observation_id":"67407c66c8ccc8ec87ccd4b6ae121aa3","value":1335.33},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"81789/50","observation_id":"c47c79a0ac9d03dc0858a3e32eb10cf4","value":1635.78},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"72223/50","observation_id":"f4a874da20b414e345395af6876bfede","value":1444.46},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"24863/25","observation_id":"9d93ac6bf71dd0301098e7d1b1db2fb3","value":994.52},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"163023/100","observation_id":"f53e9f08adf2987f0e2c59d35bf57232","value":1630.23},{"configuration":"Ling 3.1 Flash","model_id":"m-f74476b1532b3a4f","normalized_unit_value":"81319/50","observation_id":"efd709c75b89ba20347f69709bb9a3a4","value":1626.38},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"159399/100","observation_id":"d52704f212d466a4d653d53f1794d1db","value":1593.99},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"145737/100","observation_id":"727c30106a7ec7d9e1c1205d514e6937","value":1457.37},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"5459/5","observation_id":"dde8616d382b2b0a447eb1104181da5e","value":1091.8},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"40393/25","observation_id":"fcdd405fcadd197dbec16caec46c9124","value":1615.72},{"configuration":"DiffusionGemma 26B A4B","model_id":"m-fe0a26f770e1e011","normalized_unit_value":"657/2","observation_id":"4fa0d328e940e0ba1d47d7e70c6e0521","value":328.5}],"protocol_id":"e895f7b24a99582d5a133a40","unit":"分"},"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"4426376891597953/12500000000000000","observation_id":"9f3e53e10beed094c8835d4e672cc44f","value":0.35411015132783624},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"6486844882398171/10000000000000000","observation_id":"4bd5c8368df638bd0c064dd06c306d48","value":0.6486844882398171},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"3767979122295127/1000000000000000000","observation_id":"de96f31cb9e7df111014b360185583e3","value":0.003767979122295127},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"10285631346002291/500000000000000000","observation_id":"e3a197e716166eb9edc4c46f49e52295","value":0.020571262692004582},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"3008381163628863/5000000000000000","observation_id":"1b6722ff89e791988385d9c44034ffe3","value":0.6016762327257726},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"285322998928959/400000000000000","observation_id":"9d53e1ef4d2e9c319b7f0dacff98d83c","value":0.7133074973223975},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"3589329590865593/500000000000000000","observation_id":"bfaca7bcdefcec6c5d1daac93dc0d456","value":0.007178659181731186},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"14695137139561937/100000000000000000","observation_id":"eb41669f3fb02b0d8fee80ec8c840979","value":0.14695137139561937},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"7456429301855613/500000000000000000","observation_id":"427f2b62a677b7b14b79a216cd5cce2a","value":0.014912858603711226},{"configuration":"JT-4.1 Flash 236B A21B (Reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"4647120946439391/10000000000000000","observation_id":"f3bd4620346190964b5a6a49765359b6","value":0.4647120946439391},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"24071137576251893/500000000000000000","observation_id":"441205ca0d2d0b6e786d210564b45527","value":0.048142275152503786},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"14818345798643587/100000000000000000","observation_id":"ae26dd2850d89ddaa6b644b249715c9f","value":0.14818345798643587},{"configuration":"Nemotron 3.5 Bolt 74B-A9B","model_id":"m-3ad11f2b6253b364","normalized_unit_value":"309217612808113/2000000000000000","observation_id":"d9c233b856dc8ecc623629216c3b1c31","value":0.1546088064040565},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"36436554387118747/100000000000000000","observation_id":"20e66aebafb2226fab822f3a7524d54c","value":0.36436554387118747},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"18006007862431707/1000000000000000000","observation_id":"4a74a50d45014989b724a431c32f6c56","value":0.018006007862431707},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"5993009432118243/10000000000000000","observation_id":"44e1f8cce1da50e102dced9197ed9f4e","value":0.5993009432118243},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"1712293701672219/2500000000000000","observation_id":"fb60ba99063f49f876c4c8d56120f0f3","value":0.6849174806688876},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"6299648656993127/100000000000000000","observation_id":"b05e70f346089460a993097793ed0c46","value":0.06299648656993127},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"24454304615216713/100000000000000000","observation_id":"2f9ec0e2d5074f31fb106a7f0db810e6","value":0.24454304615216713},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"6405168250285657/10000000000000000","observation_id":"250157187c56354b4eed518f47426340","value":0.6405168250285657},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"4747087040091967/10000000000000000","observation_id":"6ca0b1280bb8cb8f57834c52f2472139","value":0.4747087040091967},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"916221386365139/1250000000000000","observation_id":"97fcee40bdfab50663d8d9aa18bd9131","value":0.7329771090921112},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"3404727565603879/5000000000000000","observation_id":"b2ddde3d390d97ad479920a7865821c3","value":0.6809455131207758},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"1813565262811761/40000000000000000","observation_id":"5cc05f32d32a7a675fc47d1b76ed156a","value":0.045339131570294025},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"350916867231997/125000000000000000","observation_id":"a389be9490eced3c5e8431f3790a6ecd","value":0.002807334937855976},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"23864439593997933/100000000000000000","observation_id":"94674bbf056258753d15a53bade2e692","value":0.23864439593997933},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"3277917574484729/5000000000000000","observation_id":"5dc195401c512a05556701293da42306","value":0.6555835148969458},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"3478244196003099/20000000000000000","observation_id":"1915dcb5be7d210d57854519bafaa4ef","value":0.17391220980015495},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"1568437535870999/10000000000000000","observation_id":"1aac376daa38513de1520f91177fbd00","value":0.1568437535870999},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"2891505270814033/5000000000000000","observation_id":"7fa3942a02994b01f2c4e2365b0583bc","value":0.5783010541628066},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"679669728820039/5000000000000000","observation_id":"ba2dbb617a983db0793fc0f1b006c6eb","value":0.1359339457640078},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"1039508970294839/250000000000000000","observation_id":"b4943be7200f5ac57de9f15aacbca48d","value":0.004158035881179356},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"1550269389179839/2000000000000000","observation_id":"a2113bd24aca42a48eba4972dbb6e6d4","value":0.7751346945899195},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"280952690543839/500000000000000","observation_id":"308f05669cd695ca55d1d348f97d3b83","value":0.561905381087678},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"1327884414360179/2500000000000000","observation_id":"87eb4175855e40447de186dd2f3411a6","value":0.5311537657440716},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"18104537047977383/1000000000000000000","observation_id":"c9594ae7e63800079a6fe5e87536af10","value":0.018104537047977383},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"56657663730723/78125000000000","observation_id":"b04984f7d9488e18c36d59811ea3856b","value":0.7252180957532544},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"509502447137631/250000000000000000","observation_id":"e01a487a4aa0d6109e212abc8d8c200e","value":0.002038009788550524},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"1891590945824513/125000000000000000","observation_id":"354c0d77cbcd6264333d316fdf9f676b","value":0.015132727566596104},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"7394444993005693/100000000000000000","observation_id":"7fbcd0afe5186232a5dba432e082b6ef","value":0.07394444993005693},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"742198964455803/1250000000000000","observation_id":"9a8fd23ce0ae510627402be3083ec808","value":0.5937591715646424},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"4205034817359771/500000000000000000","observation_id":"ec3633dbe4490f5d8334e3da32d471af","value":0.008410069634719542},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"3377330370758887/5000000000000000","observation_id":"115a839b24d533001f549fec19f24cc9","value":0.6754660741517774},{"configuration":"Gemma 4 12B (Reasoning)","model_id":"m-a4be17f49ae915ce","normalized_unit_value":"3427861807967129/500000000000000000","observation_id":"49cf1a20c5edfdf8e5e8bfd178db37ad","value":0.006855723615934258},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"2610203199558987/10000000000000000","observation_id":"908bb98659898bf60c101cff97ce69b8","value":0.2610203199558987},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"373293692699249/12500000000000000","observation_id":"b55d26365439e3d2ed6eba908f339027","value":0.02986349541593992},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"4984889684475201/100000000000000000","observation_id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","value":0.04984889684475201},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"1211446630516397/2500000000000000","observation_id":"981d2354449aefd49dcd7a2e96840b24","value":0.4845786522065588},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"6561018180201071/10000000000000000","observation_id":"01aaf3c58f06d7a8962fe3fd60fa4286","value":0.6561018180201071},{"configuration":"Mercury 2.5","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"11770540151643593/500000000000000000","observation_id":"43181abb2675b0ef85c7558935e2c229","value":0.023541080303287186},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"6831776269647931/100000000000000000","observation_id":"af88768655cfd944c0cc2a22a6f5661c","value":0.06831776269647931},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"17629116737896317/100000000000000000","observation_id":"c9e2b1db45e7b21cbd9a431702d99eb1","value":0.17629116737896317},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"7767473047487873/200000000000000000","observation_id":"cf48341c74514d20a48b2f428df05755","value":0.038837365237439365},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"52251479430171/250000000000000","observation_id":"b87ec7052a30aebc07b1b22c9deef3cd","value":0.209005917720684},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"311014324981321/500000000000000","observation_id":"a4b22debd5eda444528d12c717ed41a1","value":0.622028649962642},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"24121221841120963/50000000000000000","observation_id":"07fee1c32e6f2287a67ddf683ea02c91","value":0.48242443682241926},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"6664095364949257/10000000000000000","observation_id":"74476994a8c12c8ddef6748c439b2908","value":0.6664095364949257},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"3018431391083891/5000000000000000","observation_id":"e492bf3e10afd1293a1152cd005f96a3","value":0.6036862782167782},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"6953868128636069/10000000000000000","observation_id":"f100a726506576804779fdbd76208d6e","value":0.6953868128636069},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"14098964553163573/50000000000000000","observation_id":"2d900bebd8be5f37326880467236361d","value":0.28197929106327146},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"1435376425638799/2000000000000000","observation_id":"7e4dabafdeb0615d7fe8c2c4159042bd","value":0.7176882128193995},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"3681825674234479/20000000000000000","observation_id":"df8bf63c05795b45b4fd82cd57c8b504","value":0.18409128371172395},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"1860401377194007/5000000000000000","observation_id":"d30d4db6bfda99c8317f5eddf5d0e239","value":0.3720802754388014},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"2479417400536003/100000000000000000","observation_id":"3e7bd955b2670c77e55665a0183fa94c","value":0.02479417400536003},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"21251257600749407/100000000000000000","observation_id":"c964aa700a761259aa3ab0adef555cba","value":0.21251257600749407},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"1172468188683817/2000000000000000","observation_id":"45f30ef9c9577f4abce574fa0b12543a","value":0.5862340943419085},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"2485405913656597/5000000000000000","observation_id":"cc297748bec17bb0b4aa16e861980849","value":0.4970811827313194},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"121588446447251/156250000000000","observation_id":"1b1f1f3e062850814bcd282684d4e879","value":0.7781660572624064},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"69885402368961/125000000000000","observation_id":"12fa4c01c24c67c3a86dd523987aed83","value":0.559083218951688},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"212672561379327/400000000000000","observation_id":"c468f61930c57074fdae0429fd64cde2","value":0.5316814034483175},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"5002254102368191/20000000000000000","observation_id":"bdc1d1c2838fa6785d600f455b24f788","value":0.25011270511840955},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"22810660089486323/50000000000000000","observation_id":"fb9e5b876d81fca5e8cf53e8169709cb","value":0.45621320178972646},{"configuration":"Ling 3.1 Flash","model_id":"m-f74476b1532b3a4f","normalized_unit_value":"6174373352506193/10000000000000000","observation_id":"712bffe9ecc8daeca7e0dc7a1fe07e58","value":0.6174373352506193},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"159239356846759/312500000000000","observation_id":"85e95115c24e6ebd5d7a72360c176406","value":0.5095659419096288},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"2835582773172001/5000000000000000","observation_id":"63186f8028fad18deb2fa4c75b5a0888","value":0.5671165546344002},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"22252425682190993/100000000000000000","observation_id":"b1f8ac835c2c56f569ae7093d49e508f","value":0.22252425682190993},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"286236882729111/500000000000000","observation_id":"f2b687ab9797dca77a0de360cefeb5cf","value":0.572473765458222}],"protocol_id":"5434f491d6ad286fe20f125d","unit":"比例"},"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0":{"benchmark":"Terminal-Bench-Science","benchmark_version":"0.1.0","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"1/5","observation_id":"736379d5812d86c74d58e127e686a3db","value":0.2},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"580952380952381/1000000000000000","observation_id":"c9ff6ecd964f033f20c34b6056abebda","value":0.580952380952381},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"0","observation_id":"d1c19809bb6f7e59cd9274e82fb1b4b5","value":0},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"0","observation_id":"10a24da4314ddad1e5487011d9ce22be","value":0},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"144047619047619/250000000000000","observation_id":"3a64ac33a8e0bec7c9bbb47064e6200b","value":0.576190476190476},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"371428571428571/1000000000000000","observation_id":"0caa25aeefa10afc49082f57afb1971b","value":0.371428571428571},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"0","observation_id":"d7eec79b902bf9a304fdfc120e9aa124","value":0},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"0","observation_id":"b878256e74c2c14b269834f44b853cea","value":0},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"0","observation_id":"569c655345f8f6c2c6fa591358c80baa","value":0},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"0","observation_id":"42505f6b2ea9972da7c497ad966aebac","value":0},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"1/10","observation_id":"d203cb7570f8997ff9e7dbde81b27140","value":0.1},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"633333333333333/1000000000000000","observation_id":"adb8a6728c00e27c92196139aaf8ba46","value":0.633333333333333},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"0","observation_id":"1b904acff7083099b3a91e7b6c5b575a","value":0},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"428571428571429/10000000000000000","observation_id":"e9c81f27a54c74df81f07409ddb33f17","value":0.0428571428571429},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"342857142857143/1000000000000000","observation_id":"440fdb6baf103b127060630dee151d5e","value":0.342857142857143},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"257142857142857/1000000000000000","observation_id":"161dc1424fa043da938a084c4e822aa1","value":0.257142857142857},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"57142857142857/500000000000000","observation_id":"682e14ea89cac38d7b78eec67be7db59","value":0.114285714285714},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"428571428571429/10000000000000000","observation_id":"9bdac139a55bab7bdafc52c1059a96f7","value":0.0428571428571429},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"0","observation_id":"a2c09e681871381cb89f80ae44384709","value":0},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"528571428571429/1000000000000000","observation_id":"84ccb8733065be35fe6ce8fbe381732b","value":0.528571428571429},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"119047619047619/1000000000000000","observation_id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","value":0.119047619047619},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"119047619047619/12500000000000000","observation_id":"63428ab4d6f5aa0cb6c23626f378d1a8","value":0.00952380952380952},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"0","observation_id":"705b0d6767e14eda4ee6da31459437c6","value":0},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"333333333333333/10000000000000000","observation_id":"e374e482dae499c526f6a7da9baef1aa","value":0.0333333333333333},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"0","observation_id":"47e73ec519c6695efcb0b041ef5795ce","value":0},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"433333333333333/1000000000000000","observation_id":"d9f1bb911398047de48a7366176d8fef","value":0.433333333333333},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"0","observation_id":"ac28090aae70acdb086bd22103c739d0","value":0},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"428571428571429/10000000000000000","observation_id":"6b7d4617c648ad490ad35eb982d8d2e7","value":0.0428571428571429},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"0","observation_id":"13146daedf685d8da2f4bf015457ee2c","value":0},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"0","observation_id":"70737e236c952ff057602f3190f37b75","value":0},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"0","observation_id":"8120a268fd2b22fe2a65df299e1ab0c5","value":0},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"119047619047619/25000000000000000","observation_id":"c7d3f53d83f5e19ff46dc9c5f753d703","value":0.00476190476190476},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"128571428571429/1000000000000000","observation_id":"bb1f3609ad48f7662af1e4f3a15cef7a","value":0.128571428571429},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"0","observation_id":"f04dc7e5afab13c2812e16f8a7f09157","value":0},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"0","observation_id":"5d3cac863b9cb777dcf0429fca5619c7","value":0},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"0","observation_id":"d166f2e488fa71112bb60e275a7535f1","value":0},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"119047619047619/1250000000000000","observation_id":"7c5bdba52f2dfb5d75aa9320af9dd9f9","value":0.0952380952380952},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"119047619047619/25000000000000000","observation_id":"2f71320c4fe6daaec73fc9a09a2e75a6","value":0.00476190476190476},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"128571428571429/1000000000000000","observation_id":"ebff4c97ce1c3d30b95d6d855419e5c1","value":0.128571428571429},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"119047619047619/2500000000000000","observation_id":"0ca35771477057319be8d71ddeaad930","value":0.0476190476190476},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"59047619047619/100000000000000","observation_id":"8c7b6663de4bc02f5ebccae026f35f62","value":0.59047619047619},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"119047619047619/25000000000000000","observation_id":"5952fe944fd804ef75ab5466dc485304","value":0.00476190476190476},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"533333333333333/1000000000000000","observation_id":"8219d3f5fb2653720b3130ec56d5201c","value":0.533333333333333},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"0","observation_id":"2f6e14dbb9becc8bf0feb78f2973083f","value":0},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"119047619047619/25000000000000000","observation_id":"1a6a33cd4cc454e60378e0a0694263f3","value":0.00476190476190476},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"571428571428571/10000000000000000","observation_id":"c254195ad65c522fa9609ac4a64d415d","value":0.0571428571428571},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"19047619047619/1000000000000000","observation_id":"356d1ac12dc91c2d1cadb615b7e3023b","value":0.019047619047619},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"80952380952381/125000000000000","observation_id":"416c079f92e7711d76f35d3a6da3f408","value":0.647619047619048},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"857142857142857/10000000000000000","observation_id":"6de4c10b838de9bb43446d62c1dedc6c","value":0.0857142857142857},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"0","observation_id":"2c2a68b20b30379e3063400c424fa6d0","value":0},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"571428571428571/10000000000000000","observation_id":"7ead6883bce4c045cf26f5b116b4cd48","value":0.0571428571428571},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"119047619047619/5000000000000000","observation_id":"2b8cf679ed4d4658129609f163b9f607","value":0.0238095238095238},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"571428571428571/10000000000000000","observation_id":"f1eefd5ab642da9d01de8eac565cc324","value":0.0571428571428571}],"protocol_id":"a09e264ffddc775882a3b90f","unit":"比例"},"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e":{"benchmark":"Humanity Sixth Sense","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6-Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"30","observation_id":"8a450f924e67dc0b617d0ac58b1154ec","value":30},{"configuration":"GPT-6.1-Sol (max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"233/5","observation_id":"3a0983451c53206a0fa5061a863b8a25","value":46.6},{"configuration":"Gemini-3.7-Flash (high)","model_id":"m-16a8dce2a6e7167a","normalized_unit_value":"199/5","observation_id":"e0d9bd2a6ca55a94058a8fdfafb19d29","value":39.8},{"configuration":"Claude-Sonnet-5 (max)","model_id":"m-2cd37409a483435a","normalized_unit_value":"124/5","observation_id":"cf4a733238c901c9de673ff40841fce0","value":24.8},{"configuration":"Muse Spark 1.2 (xhigh)","model_id":"m-2f22b07da3af863c","normalized_unit_value":"49/2","observation_id":"b1617164f055d3d72ddaf46f34da89d8","value":24.5},{"configuration":"GPT-6-Sol (max)","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"156/5","observation_id":"e25930e774ad7abbc0f5c7ffcc567aa5","value":31.2},{"configuration":"Gemini-3.8-Flash (high)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"208/5","observation_id":"11851315d0cbcd3709f1ea2206b9b74b","value":41.6},{"configuration":"GPT-6-Astra (max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"268/5","observation_id":"35c55ac70a4db5208200b7b35fe696c9","value":53.6},{"configuration":"Claude-Opus-5 (max)","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"61/2","observation_id":"175d0f5744557d89454896c1420ebf13","value":30.5},{"configuration":"MiMo-v2.6-Flash (high)","model_id":"m-60cd64beda366768","normalized_unit_value":"127/5","observation_id":"672a5ff05bade6e76dd2270c17366a35","value":25.4},{"configuration":"GPT-5.6-Luna (max)","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"108/5","observation_id":"6698824b06580d547cc03c11c24e9e05","value":21.6},{"configuration":"Muse Wonder 1.0 (max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"437/10","observation_id":"b9ee5aa138e25de06492fbd597ad313a","value":43.7},{"configuration":"Gemini-3.6-Flash (high)","model_id":"m-77b5dbc317075e9d","normalized_unit_value":"319/10","observation_id":"6d9db035c8e86692a9eb736f0a2c02ca","value":31.9},{"configuration":"Gemini-4-Argon (high)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"277/5","observation_id":"18175febe04f7246f991614fc6ca93f9","value":55.4},{"configuration":"Claude-Fable-5 (max)","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"69/2","observation_id":"cf924b68bd130c394ca909ce8970cb90","value":34.5},{"configuration":"Gemini-3.5-Flash (high)","model_id":"m-8d99798a7e24690c","normalized_unit_value":"163/5","observation_id":"1fedb85017624b14b5d56927cae80633","value":32.6},{"configuration":"Claude-Fable-5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"204/5","observation_id":"2d80b1607b40fc4d214cb22e259d1221","value":40.8},{"configuration":"Qwen3.8-Max (max)","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"33","observation_id":"b0de0689b04f1b90ca42e94e49e0a3bb","value":33},{"configuration":"Kimi-K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"51/2","observation_id":"d100a26c5dab08a99d576ce25ec3994e","value":25.5},{"configuration":"Muse Spark 1.3 (max)","model_id":"m-c0580b01221ed260","normalized_unit_value":"187/5","observation_id":"fcb2b9f9c2c7859d3369fef86db555f4","value":37.4},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"301/10","observation_id":"60e8851ddeb919672b146036d9d51182","value":30.1},{"configuration":"Claude-Opus-4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"47/2","observation_id":"2c5710356bbeeb960760bc793dbd6d80","value":23.5},{"configuration":"GLM-5.3-Flash (max)","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"134/5","observation_id":"7d2a87235efadfc965d04473b687dd29","value":26.8},{"configuration":"Claude-Opus-5.5 (xhigh)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"223/5","observation_id":"78db71948ffd6a9f7a24a59779294d90","value":44.6},{"configuration":"MiMo-v2.6-Pro (high)","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"126/5","observation_id":"73994f722364016b18f62a10e6695b38","value":25.2},{"configuration":"Muse Spark Open (max)","model_id":"m-db208dec31031420","normalized_unit_value":"148/5","observation_id":"fe8e2c3470907e5adda03a06d7b5f363","value":29.6},{"configuration":"GPT-6-Luna (max)","model_id":"m-e5d076977e951607","normalized_unit_value":"21","observation_id":"c23775cd50aa5f64341fa34e8b1cf6d3","value":21},{"configuration":"GPT-5.6-Terra (max)","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"227/10","observation_id":"2d9aead127b28a667d7f4b6c145f9945","value":22.7}],"protocol_id":"e86182c6397db78c5990f7ac","unit":"%"},"16f82f43e67c371ae3dcb6f2e0649c38989e5cd62f81e3b9e3b31df741787aa6":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"559/10","observation_id":"130da9dac3f554ec6bf0afc38a355782","value":55.9},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"513/10","observation_id":"0cba5bb1737ffc1f4e69a7db0bd621bc","value":51.3},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"351/5","observation_id":"13ab52d0b216c84bb8c4111e06baab4e","value":70.2},{"configuration":"Solar Open 2 250B-A15B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"863/10","observation_id":"24b5b69a17fb0e9c1efc24d9973f4c36","value":86.3},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"279/5","observation_id":"c750a2ef43ab55535245679f96847a15","value":55.8}],"protocol_id":"62f4969756e5cb55e0903043","unit":"%"},"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0":{"benchmark":"CritPt","benchmark_version":"2 Hard","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"87/200","observation_id":"b0085122202e14e07e4a784b0954985f","value":0.435},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"139/200","observation_id":"96d62eb6c5d675afc73f8103f6dbc627","value":0.695},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"0","observation_id":"e23c22def1e7c8613f7559601fcfa581","value":0},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"0","observation_id":"8e1dd162a62fd85fef38ef1d97e4e9bc","value":0},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"17/25","observation_id":"624fa427f469bd9c8115ca8491b3f8a8","value":0.68},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"113/200","observation_id":"f2c7d3b6f687d90b18b5ecae64664442","value":0.565},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"1/100","observation_id":"9fef9d533791c4fe2ca276032af2b566","value":0.01},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"1/100","observation_id":"1ce53974b2fd41332a137498132cd83b","value":0.01},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"37/200","observation_id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","value":0.185},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"27/200","observation_id":"e6e2851045fbef435d3da74124d06523","value":0.135},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"93/200","observation_id":"4e088fc1433335f9f212c9f9ffa8b974","value":0.465},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"141/200","observation_id":"0a974b1036bafb40411da54c5245579a","value":0.705},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"1/200","observation_id":"98267d8baf13117a13191a66537ee6eb","value":0.005},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"59/200","observation_id":"737809004b185d439bc3bdb639d99a34","value":0.295},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"99/200","observation_id":"e4f3938369b9a55c0c1841556cc18184","value":0.495},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"101/200","observation_id":"024c0dcfccf03a6bb518b6c682477674","value":0.505},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"67/200","observation_id":"539eef88fa72b976d41f8f0e6b20d640","value":0.335},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"27/100","observation_id":"e52491cdc38740f220e3951458e0d11e","value":0.27},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"9/50","observation_id":"7de1f456199243fb6a03dcf75e8fd06c","value":0.18},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"21/40","observation_id":"3130f94996ac2afcc2f6ea8e78d8c87e","value":0.525},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"59/200","observation_id":"b1d4b1cfdd4d38290545201e6c39db63","value":0.295},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"31/200","observation_id":"2147a64a171deda69e1dcacdea28ad28","value":0.155},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"1/40","observation_id":"bfeb15a230a3fe93a4b8668ae5c71e33","value":0.025},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"13/40","observation_id":"321a79f217d130c0ceb39102c3c1fab9","value":0.325},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"3/200","observation_id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","value":0.015},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"21/40","observation_id":"2a25d966a3bebfa161181074ba7fbad0","value":0.525},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"3/200","observation_id":"2f96f5af894d02c94d57177465def23b","value":0.015},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"77/200","observation_id":"20d35e57ebd245b7509c38232fe59998","value":0.385},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"2/25","observation_id":"d6e7cb95527cd29a8925ca94f594e568","value":0.08},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"3/20","observation_id":"f5fbcfce1c679af1818579e2368d3d27","value":0.15},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"9/50","observation_id":"a9467dab432c3fd1ab9b333380c66f83","value":0.18},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"53/200","observation_id":"bd76f518c9968d380eda9d90a689c8c2","value":0.265},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"39/100","observation_id":"6c2b0cbfd16faf311bd4edca68e40d0e","value":0.39},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"7/100","observation_id":"142184a428fc22f5f377c4fcd96f6a10","value":0.07},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"1/25","observation_id":"55904790fd4c7a2e974218f474b3d773","value":0.04},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"3/200","observation_id":"3441fc8b1a39080bc19e5b920fef2c87","value":0.015},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"67/200","observation_id":"6177f516623a8dfc4d1c35316f9c20f2","value":0.335},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"1/25","observation_id":"a3193e796745f7cc574aae69daceb5f1","value":0.04},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"1/4","observation_id":"d9b68c289d550eec511fa758a53ee05a","value":0.25},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"49/200","observation_id":"620b6f3e6314727d16068f5503869955","value":0.245},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"17/25","observation_id":"46e4906d019b793c19b1d443cd35b178","value":0.68},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"59/100","observation_id":"57f29ac78c3749b38fa2eef65ef43c04","value":0.59},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"3/25","observation_id":"5f0a8e1454688084cb0e096162e1eff2","value":0.12},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"7/25","observation_id":"3d5749511482d0fbdaddfa896055ac6a","value":0.28},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"17/50","observation_id":"2f0eebfa718074ef0cab557ff07d209b","value":0.34},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"59/200","observation_id":"3eef34fe8a68f3f65cb23f07be80c5e4","value":0.295},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"69/100","observation_id":"29eed13fef0ca4e5917825122c33b6b0","value":0.69},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"17/40","observation_id":"a69315d97d8db2867eda5787e5741424","value":0.425},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"7/200","observation_id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","value":0.035},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"7/20","observation_id":"34dffe7c2a7d437da340dfb5246fe9c8","value":0.35},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"13/40","observation_id":"b5e83ad92c83f666dbcd3cdc392d63b6","value":0.325},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"47/200","observation_id":"a2d7db240ac12326c1a9bfd85ff9b349","value":0.235}],"protocol_id":"ad076bfd835af583f4130282","unit":"比例"},"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"98/5","observation_id":"da2362332ab0039d90ab52612d3c3148","value":19.6},{"configuration":"GPT-6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"207/5","observation_id":"03836c4b9ecec508c6c4b2750f3aead4","value":41.4},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"269/10","observation_id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","value":26.9},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"171/10","observation_id":"8ad20dc033b0a09d811fcebfca51c2ab","value":17.1},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"157/5","observation_id":"a5d016c51f0b0d94dac22663791c4a6b","value":31.4},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"40","observation_id":"b4428690505427f75d65ee7f08fbf5db","value":40.0}],"protocol_id":"659066a5e56ba232a06f7c9f","unit":"%"},"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830":{"benchmark":"OSWorld","benchmark_version":"2.1","members":[{"configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"809/10","observation_id":"2ed0ffc858d7ba150a21822c243ebff1","value":80.9},{"configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"747/10","observation_id":"f61750e8ba74bd2c7928dd7c56124cb4","value":74.7},{"configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","model_id":"m-98ee7005cf45b532","normalized_unit_value":"384/5","observation_id":"84a17f8a2fe109b449e2d00efd604380","value":76.8},{"configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","model_id":"m-cfe0652e9d685783","normalized_unit_value":"436/5","observation_id":"c81051296c04b2de65d68a3054256560","value":87.2},{"configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"163/2","observation_id":"f7b8bce78b6bbb414e44451ecece6456","value":81.5}],"protocol_id":"ab44942b390af73e564b6930","unit":"%"},"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f":{"benchmark":"IFBench","benchmark_version":"current published task suite","members":[{"configuration":"MiniCPM-V 4.6 1.3B","model_id":"m-0cd6b9e2d04f2c30","normalized_unit_value":"26734693877551/100000000000000","observation_id":"e359de5ff096ac45c506a2baeb714ed7","value":0.26734693877551},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"278231292517007/500000000000000","observation_id":"950f9e2f9181b2c99ee5f8151838596d","value":0.556462585034014},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"445578231292517/1000000000000000","observation_id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","value":0.445578231292517},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"575510204081633/1000000000000000","observation_id":"65c9161268988be090ca4411568e9e7b","value":0.575510204081633},{"configuration":"JT-35B-Flash","model_id":"m-2d9c4a1754ad2072","normalized_unit_value":"210204081632653/500000000000000","observation_id":"0ada415551e6be20a6063101be020eb6","value":0.420408163265306},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"42984693877551/62500000000000","observation_id":"5fbfa4015dcb4ab37c1080316055d055","value":0.687755102040816},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"631292517006803/1000000000000000","observation_id":"15ec6c548623aedd9b87ffa73167943d","value":0.631292517006803},{"configuration":"Nemotron 3.5 Bolt E4B 4GB (Reasoning)","model_id":"m-6d1e332cb8eb6a77","normalized_unit_value":"610884353741497/1000000000000000","observation_id":"8b8b563792b7986f8d696ab64682e278","value":0.610884353741497},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"389795918367347/500000000000000","observation_id":"c4485d942b9b21c7a152b1888b11b7c4","value":0.779591836734694},{"configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","model_id":"m-7b2a8fe101678ebd","normalized_unit_value":"83078231292517/125000000000000","observation_id":"ea00f82eb447a348674a4f2b40a4a4f4","value":0.664625850340136},{"configuration":"Command A","model_id":"m-8877a9675c79fcad","normalized_unit_value":"45578231292517/125000000000000","observation_id":"3f2a6e6ace968612fc72d2719d04f5ee","value":0.364625850340136},{"configuration":"Gemma 4 12B (Reasoning)","model_id":"m-a4be17f49ae915ce","normalized_unit_value":"91921768707483/125000000000000","observation_id":"af2b7a8fe076debb535a31ef7b09244c","value":0.735374149659864},{"configuration":"MiniCPM5-1B (Reasoning)","model_id":"m-ab31824c5d59e70f","normalized_unit_value":"98639455782313/200000000000000","observation_id":"02353ea61ea4b75a5ac63d23982f65dc","value":0.493197278911565},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"813605442176871/1000000000000000","observation_id":"74e2baf55ca27065ec7aeb8f6eee9414","value":0.813605442176871},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"330952380952381/500000000000000","observation_id":"353142d52584e4a6aac9d52541e06d23","value":0.661904761904762},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"828571428571429/1000000000000000","observation_id":"89b2619aee2f39a13fb5f1677cbe7c28","value":0.828571428571429},{"configuration":"DiffusionGemma 26B A4B","model_id":"m-fe0a26f770e1e011","normalized_unit_value":"148639455782313/250000000000000","observation_id":"8e2367487c04d7937c7f24678241f073","value":0.594557823129252}],"protocol_id":"7b8df779a6b066873d2ea56f","unit":"比例"},"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"GPT5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"941/10","observation_id":"3d74a21fb4db168c492ba73576bd0b2e","value":94.1},{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"903/10","observation_id":"4bee782a627e367cbb2995c9ffc85564","value":90.3},{"configuration":"Fable5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"463/5","observation_id":"d42da14e4f5c6f46bc80f80cc52a8e79","value":92.6},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"463/5","observation_id":"de593678281918b51aaf95adefe186dd","value":92.6},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"462/5","observation_id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","value":92.4},{"configuration":"Opus4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"92","observation_id":"c316ffaf81b4d0057dc56f75f0232808","value":92.0}],"protocol_id":"8ef2772079cc631a5830d286","unit":"%"},"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8":{"benchmark":"AA-Omniscience","benchmark_version":"1.1","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"6576530612244897/10000000000000000","observation_id":"b11d60a588bf725879871b88bc8f70cb","value":0.6576530612244897},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"1924890829694323/4000000000000000","observation_id":"8c01a1619e7c484574ab42025543f92a","value":0.48122270742358075},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"6967694566813509/10000000000000000","observation_id":"7f11b1b4129f0979e12158aab3657dcc","value":0.6967694566813509},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"6777483944087647/10000000000000000","observation_id":"5f020104c70b0364755cf0f561bff098","value":0.6777483944087647},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"6549968173138129/10000000000000000","observation_id":"5ac877348d0fb5ad33d4ed61b4154929","value":0.6549968173138129},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"4121964991530209/5000000000000000","observation_id":"211e11b4d985ead9c997be1fc30d7961","value":0.8243929983060418},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"154228855721393/1000000000000000","observation_id":"2c99a44c194b274fedb14bdd3896ace9","value":0.154228855721393},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"16496529195590037/100000000000000000","observation_id":"30b64e451b40a0bb848f166f0f44b6dd","value":0.16496529195590037},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"3949169110459433/10000000000000000","observation_id":"01245c922f864bcfcfa4ad1336dea016","value":0.3949169110459433},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"6951374207188161/10000000000000000","observation_id":"1c32657b7924b6111d38086436775d4a","value":0.6951374207188161},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"11219676549865229/25000000000000000","observation_id":"4b5e9db335123a5e12d2351e281e6182","value":0.44878706199460916},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"2510841283607979/5000000000000000","observation_id":"6e1af818db0c40a97c73d45f3dd92647","value":0.5021682567215958},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"9896058899956689/50000000000000000","observation_id":"d7d18670af809e53c710911ecd277233","value":0.19792117799913378},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"47756410256410253/100000000000000000","observation_id":"b4e27389bf022ba2e7fea370cfc23e71","value":0.47756410256410253},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"97911227154047/156250000000000","observation_id":"faf92c6808beb316ccdbea4a34863bce","value":0.6266318537859008},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"7912856374394837/10000000000000000","observation_id":"58f09829d0f39ffd8664ffd449b9bfbd","value":0.7912856374394837},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"13319011815252413/100000000000000000","observation_id":"ff2df3ac7046d17ae8e64e925f03f746","value":0.13319011815252413},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"7998836532867947/10000000000000000","observation_id":"5789f26e6b16a0b75b183dcbb8cf44bd","value":0.7998836532867947},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"359009009009009/500000000000000","observation_id":"8175922707070fb09367c5a6424e1c3a","value":0.718018018018018},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"1581528402125051/2000000000000000","observation_id":"9364a9aa01cea849d2eb414760bb16ec","value":0.7907642010625255},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"9574468085106383/10000000000000000","observation_id":"2f9d0f30b6d18925324be8097217f20f","value":0.9574468085106383},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"911778618732261/1250000000000000","observation_id":"602505b7f91e3456e6da6058f7b120bf","value":0.7294228949858088},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"4090566037735849/5000000000000000","observation_id":"d5ec18e90fffd664c6089d00e270f823","value":0.8181132075471698},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"68909426987061/80000000000000","observation_id":"862933c1b9c8be1d0cbb4bb130b222e9","value":0.8613678373382625},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"6310679611650483/100000000000000000","observation_id":"cd0ef2b66ec6818e014beedc4ad88c9f","value":0.06310679611650483},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"1833458505444987/2500000000000000","observation_id":"82eab72497730b1d30475a9ca4c7819b","value":0.7333834021779948},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"7489451476793249/25000000000000000","observation_id":"2cbd1e01e64dad121dca7aaf300f12be","value":0.29957805907172996},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"1607701329163409/2500000000000000","observation_id":"2184077575074c69e38c71abae562f14","value":0.6430805316653636},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"6370235934664247/10000000000000000","observation_id":"ae54534276a6c4c19dcb828d96fdee4c","value":0.6370235934664247},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"154803216250529/200000000000000","observation_id":"f659a47d7f8491b1073ef450bc951aa0","value":0.774016081252645},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"3278418451400329/10000000000000000","observation_id":"3698e24fd77dd4f8bc27edd44ed2efc7","value":0.3278418451400329},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"5899903753609239/10000000000000000","observation_id":"111f3e2d8b49ec87b3bbe30f22c6238c","value":0.5899903753609239},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"2731512325116589/25000000000000000","observation_id":"7d74696c1e12470b21a043973afd962c","value":0.10926049300466356},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"410689495603887/2500000000000000","observation_id":"934c0d476e0913dedfc11ab20974d89e","value":0.1642757982415548},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"6962962962962963/10000000000000000","observation_id":"3bcf44016a4447d1427e1bf692c1e1bc","value":0.6962962962962963},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"3056650246305419/5000000000000000","observation_id":"330a13ca3aada5b9674ba75668e0acfc","value":0.6113300492610838},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"3569125352680371/5000000000000000","observation_id":"ee80bfee31969adf86abcbeaae10d4e4","value":0.7138250705360742},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"7185430463576159/20000000000000000","observation_id":"cd12ad2957c69fc5e8660a959eaae5db","value":0.35927152317880795},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"6165273909006499/10000000000000000","observation_id":"7e86ee60903a41ed29953982d04673a1","value":0.6165273909006499},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"4966183574879227/10000000000000000","observation_id":"035a4d676bedc8d21713e0b6aa5bc0ad","value":0.4966183574879227},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"5673076923076923/10000000000000000","observation_id":"553708168125ea365f4f1a89df474d93","value":0.5673076923076923},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"3547726341663253/5000000000000000","observation_id":"7d19cde6db2f8d3d5e0cb451c988a92a","value":0.7095452683326506},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"8645707376058043/100000000000000000","observation_id":"e14051d6dbdd6fd61ec4748398bf8936","value":0.08645707376058043},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"4971881390593047/12500000000000000","observation_id":"b3118336d86ca34d1fd6619d5a595882","value":0.39775051124744376},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"1799884991374353/2500000000000000","observation_id":"0b6767461c06c4b95e61401653221d25","value":0.7199539965497412},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"8857827476038339/10000000000000000","observation_id":"91da142d2c0422108afa80f33bfb9bb4","value":0.8857827476038339},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"532258064516129/1000000000000000","observation_id":"8283a5fc47f30f74b98c14b10fde0469","value":0.532258064516129},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"10325786858089453/50000000000000000","observation_id":"7ec06e89d9178a8c5c27508ed35805db","value":0.20651573716178906},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"7328798185941043/10000000000000000","observation_id":"4347afed89affe307401c55cb1f477b8","value":0.7328798185941043},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"1466532460996477/2000000000000000","observation_id":"1a5d81b92d3a8437744af4b44b483171","value":0.7332662304982385},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"858433734939759/1250000000000000","observation_id":"880d1169acb398e66460aed2eb721de0","value":0.6867469879518072},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"5900243309002429/100000000000000000","observation_id":"775332c591df3bb4219419f088bb5ff5","value":0.05900243309002429},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"6347329800244599/10000000000000000","observation_id":"973d182535a01b720897f5a733a8d1b0","value":0.6347329800244599},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"409681729834791/625000000000000","observation_id":"7f66b653a79b36bb8c9fbcd69744e5f5","value":0.6554907677356656}],"protocol_id":"57ce78dfdf41b4d0acd6e37d","unit":"比例"},"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"727/10","observation_id":"ff5c1b3b8cf739706c1637caa41bfd3b","value":72.7},{"configuration":"Gemini 3.8 Flash","model_id":"m-422e5ba87859dd55","normalized_unit_value":"369/5","observation_id":"e589efcb9de8d6dea5758589d780110f","value":73.8},{"configuration":"GPT‑6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"741/10","observation_id":"948d62640ac1b6f74166a175a86cf6a7","value":74.1},{"configuration":"Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"737/10","observation_id":"d36cdba4dae45a238d9898e2ce34c241","value":73.7},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"336/5","observation_id":"b632e7c137b19cc34931e9481ad27bf9","value":67.2},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"699/10","observation_id":"0772cd4ebc7a33c01cf749b6a69bd7c0","value":69.9},{"configuration":"Claude Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"337/5","observation_id":"a9df38e2550ef9f8d566441fba3e6c66","value":67.4},{"configuration":"Claude Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"59","observation_id":"dee93fd91e74e8ba6728dba7ca4fd982","value":59.0},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"348/5","observation_id":"99e7d048583a82e4719e8e299872ac27","value":69.6}],"protocol_id":"1b6f5b9c917154d5a53d4023","unit":"%"},"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c":{"benchmark":"IFBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"727/10","observation_id":"3cf0fe91d3392b3852bb737856d256d6","value":72.7},{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"791/10","observation_id":"9c0623cc90856dd0639a89861f73ff03","value":79.1},{"configuration":"Fable5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"127/2","observation_id":"3f5e26906c244a4d4f9e57725da3d9df","value":63.5},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"414/5","observation_id":"57da33d0598ebe148085f216c7c7cefb","value":82.8},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"791/10","observation_id":"1be1bd504f0391002f7de4a46514d7c5","value":79.1},{"configuration":"Opus4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"311/5","observation_id":"6d5a13969cd11982fa7a1306707181d9","value":62.2}],"protocol_id":"af6f81a701fa5f1daec85fee","unit":"%"},"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02":{"benchmark":"BFCL","benchmark_version":"4","members":[{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"97/2","observation_id":"602ba59a6dfe14df55137c17751e7666","value":48.5},{"configuration":"LFM2.5-2.6B (2.6B)","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"1422/25","observation_id":"78759a40401be7663e3957b80c649826","value":56.88}],"protocol_id":"5c8ad28d0b22630243378008","unit":"%"},"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"腾讯官方比较表 · GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"947/10","observation_id":"d009cc9a0555aee5404869255b905141","value":94.7},{"configuration":"腾讯官方比较表 · Hy4 preview","model_id":"m-383f40728e3e5259","normalized_unit_value":"923/10","observation_id":"11f1be50c8d04272ce6fde6b77e7d32d","value":92.3},{"configuration":"腾讯官方比较表 · Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"933/10","observation_id":"b2df261aa6e00c4606b5fc3f6b802f0f","value":93.3},{"configuration":"腾讯官方比较表 · Qwen3.8 Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"461/5","observation_id":"9c1d87da906e98d13d0eb89f034d7898","value":92.2},{"configuration":"腾讯官方比较表 · Kimi K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"464/5","observation_id":"d4511580b7749fdce0eb519414c56c16","value":92.8},{"configuration":"腾讯官方比较表 · GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"457/5","observation_id":"7bacdcf72f2524d7ecbbadf7b0f6e174","value":91.4},{"configuration":"腾讯官方比较表 · Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"909/10","observation_id":"8a7def0b05a244600845308f3acf91c6","value":90.9},{"configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","model_id":"m-f936349cbdb1def9","normalized_unit_value":"917/10","observation_id":"6f23f1e17caba14beed3e77ae9ea4b43","value":91.7}],"protocol_id":"8ac3a853c44f6c551fe64a3e","unit":"%"},"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"DeepSeek-V4-Flash-0731","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"272/5","observation_id":"72c84440fb70eb9d01c3241fc4d3e551","value":54.4},{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"33/2","observation_id":"9bf5f2dc87f09a1babb07763838a2e08","value":16.5},{"configuration":"Qwen3.8-27B","model_id":"m-c947385433db3f03","normalized_unit_value":"211/5","observation_id":"b993c11069821aca78f26acbce738dbc","value":42.2},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"587/10","observation_id":"80a064d01bdbcc1f636799e9dd23518a","value":58.7}],"protocol_id":"de8e60f94c1bc38e345e7379","unit":"%"},"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"73","observation_id":"8fac1a5b5e9d5b95efd26212af5182b5","value":73.0},{"configuration":"DS-V4-Flash","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"272/5","observation_id":"5aa8713728b0b23d443020c1571a6c86","value":54.4},{"configuration":"Opus-5.0","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"74","observation_id":"4fbfacefe470a5ba3c84b133b49ffefd","value":74.0},{"configuration":"DS-V4.1-Flash","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"371/5","observation_id":"3bd3c3d2983a2fb1db83fc1872674955","value":74.2},{"configuration":"K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"135/2","observation_id":"175facc5a97ac64a62324f8687068b9f","value":67.5},{"configuration":"GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"669/10","observation_id":"bbb96b1ec88d510cf9f1da8d736fb7f4","value":66.9},{"configuration":"DS-V4-Pro","model_id":"m-f936349cbdb1def9","normalized_unit_value":"627/10","observation_id":"63840894bc67cdd6c5a9cd6e26e8b9d7","value":62.7}],"protocol_id":"b229959842eb7c98b68856d4","unit":"%"},"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7":{"benchmark":"GDP.pdf","benchmark_version":"current published task suite","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"307/10","observation_id":"5f7d8e688eab4e132d7071fdcec04c3f","value":30.7},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"227/10","observation_id":"706d4a87bdc7adf93c73d7e4099021c8","value":22.7},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"149/5","observation_id":"134a991b936b3bccddf2a9a60c450e69","value":29.8},{"configuration":"Claude Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"45/2","observation_id":"dc77f0ff3c077230aef1b2b8cb04ec92","value":22.5},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"247/10","observation_id":"2f6ca6bd54419e351a2ba1a4c64dedd7","value":24.7}],"protocol_id":"7ca6f7acd460b88b1d816a01","unit":"%"},"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1":{"benchmark":"GDP.pdf","benchmark_version":"current published task suite","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"26/125","observation_id":"a183b88682c41fc78cc2ad5e5485bbb7","value":0.208},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"31/100","observation_id":"f0c0c3073a678bbcb3f5f855da9adff8","value":0.31},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"1/500","observation_id":"d9c7d28e2d9651e546a84191d6ab8ec7","value":0.002},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"3/500","observation_id":"8eaabce7efa57428ec82adf89ba50dbc","value":0.006},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"143/500","observation_id":"5a11c32835ae26c4dc93dd63135b1a2a","value":0.286},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"163/500","observation_id":"b8e5730fd93ac21867d62c37b5d60e40","value":0.326},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"19/500","observation_id":"5ecbc58b2efe8e4c7ad00b00fd409205","value":0.038},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"4/125","observation_id":"e8733c3c5f23f9a2dee5ae8306fe34a6","value":0.032},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"9/500","observation_id":"96fc6325ca8f16276f80447a5bb8a745","value":0.018},{"configuration":"JT-4.1 Flash 236B A21B (Reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"43/500","observation_id":"c038eca51adef7673b4117a8c612883d","value":0.086},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"14/125","observation_id":"17242a9341b3cc9173ea459907ce9455","value":0.112},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"2/25","observation_id":"26277d899631407b16323e0f427ca69b","value":0.08},{"configuration":"Nemotron 3.5 Bolt 74B-A9B","model_id":"m-3ad11f2b6253b364","normalized_unit_value":"7/250","observation_id":"b6eecb655a01ff6addbb4005e50d5c51","value":0.028},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"101/500","observation_id":"9ab89470db83472d1b5d2ed28ebc6783","value":0.202},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"7/500","observation_id":"9ffb2abc9760a5a8e507be89d245bb8f","value":0.014},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"21/100","observation_id":"1bf434a1cea1439ab199e26911f14113","value":0.21},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"31/100","observation_id":"96f5c4d98f214c070d53868688e0a1a3","value":0.31},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"7/250","observation_id":"2bd6dbb1c5200680f554dfce1a26feb0","value":0.028},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"14/125","observation_id":"60b6aea7e19a90a921731bc7d92e251c","value":0.112},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"9/100","observation_id":"0f9316648f89e3d521fa219b6361fd6c","value":0.09},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"3/25","observation_id":"b7c40d3349d6abfac4e0b93abf2adfa2","value":0.12},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"157/500","observation_id":"24d56f9893e1402bb483f6bef166ba08","value":0.314},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"153/500","observation_id":"a4934e52e2688102b1f21d91ca098935","value":0.306},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"8/125","observation_id":"c1a1527152276d41667adbe5c22b811f","value":0.064},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"1/125","observation_id":"7738f3afb94846ed2b13073b7d2cb3b1","value":0.008},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"27/500","observation_id":"3fef1cbee3ec381cc5e40ecf79f229ab","value":0.054},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"1/5","observation_id":"06f207d047d6194d347719cbe681b11e","value":0.2},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"61/500","observation_id":"850da6d9475999744f930f21d7861dd9","value":0.122},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"21/250","observation_id":"8a1843eddbf9ae3836989a362277ba86","value":0.084},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"22/125","observation_id":"a6443ea346efcda5f00f1ce182e0d982","value":0.176},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"7/100","observation_id":"75cb1248936655503fa568c6e5bc1e02","value":0.07},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"3/250","observation_id":"d70ece05a7cd9206504e0831b954b698","value":0.012},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"109/500","observation_id":"da4cc7b10c537dba69aca5448e60f8f7","value":0.218},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"57/250","observation_id":"5928f02dfd05733dd4d23f64991980e9","value":0.228},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"14/125","observation_id":"632aef0c9b2c1ee60a8fee169a7ead99","value":0.112},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"1/100","observation_id":"6612c73a65ff838439bc02c707904f4b","value":0.01},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"26/125","observation_id":"96a35bebfd1cf232725317d8f36f4c53","value":0.208},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"1/125","observation_id":"b9fa019c4e1362fc93bc4088a9dc8dbc","value":0.008},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"9/500","observation_id":"6e95a209bd026922387753b507be18e1","value":0.018},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"27/500","observation_id":"0e26cf9d2beafb47ed1ed89556040c14","value":0.054},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"131/500","observation_id":"5d9e8383609a67f5871ab595963f24e2","value":0.262},{"configuration":"Motif 3","model_id":"m-9e1e0f605d47cf02","normalized_unit_value":"13/250","observation_id":"b0db480f387627aef52b7f9b3d6804ac","value":0.052},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"3/100","observation_id":"40a272cc2b1879bb5b0104ebb8833d18","value":0.03},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"63/500","observation_id":"d7a0ab07d52625e0d345f1bbf275edda","value":0.126},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"21/500","observation_id":"14efb36fead599ff0b956fa101f23a86","value":0.042},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"1/20","observation_id":"994833badab5af26add99f20244353b3","value":0.05},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"16/125","observation_id":"af7da6ed00064e0d491020c30636379c","value":0.128},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"29/250","observation_id":"25ac15cd539fc43d4a096d3dea9f7966","value":0.116},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"27/125","observation_id":"b84d512b0d7ca681d5568eba0c2a9675","value":0.216},{"configuration":"Mercury 2.5","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"9/250","observation_id":"8f8d77f644fefbdcb7d35712e527c726","value":0.036},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"1/10","observation_id":"0849343c6ca0b1de586bda128e33678b","value":0.1},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"63/500","observation_id":"312a7fa68250f3fcf7a32404595e6dc4","value":0.126},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"21/500","observation_id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","value":0.042},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"4/125","observation_id":"48f1c6876fabfb820ee570e743c9611a","value":0.032},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"14/125","observation_id":"ba325e0068d64ba19ce176fe95dbb3e4","value":0.112},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"41/250","observation_id":"4189c00054cffddacad0958de8d0f156","value":0.164},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"43/250","observation_id":"b06f9a392143fecd0d26f4f434ebd403","value":0.172},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"77/500","observation_id":"f95dbe30b012b632cab81f4598931983","value":0.154},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"131/500","observation_id":"77cd16d89e49025a3285c182b2da553d","value":0.262},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"14/125","observation_id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","value":0.112},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"129/500","observation_id":"7992d55958a238d712427a450a500af9","value":0.258},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"14/125","observation_id":"2081b40577b770d6210bb3099918429f","value":0.112},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"37/500","observation_id":"65f6a64cea971ea25fa466fbee3340a6","value":0.074},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"8/125","observation_id":"10af516d49f62278a451fea09c7878c1","value":0.064},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"49/500","observation_id":"7e0fa158109c3646bd13a80e33ec79ee","value":0.098},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"24/125","observation_id":"0c01c6860fffc32ef81ab819ece21423","value":0.192},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"93/500","observation_id":"7048e36b711cfdf67298d82bfa3d5f86","value":0.186},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"71/250","observation_id":"2ef7660125e00e7ecf9ebbabeb38c644","value":0.284},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"39/250","observation_id":"046ca38af4c5ad3c8241882797259027","value":0.156},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"57/250","observation_id":"ff1f819163bf03455ec9cbe6d10872e3","value":0.228},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"17/125","observation_id":"b1d8307905906d9eadee1409ffb734e6","value":0.136},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"63/250","observation_id":"792f8618c792ce48ee9c7c6155488bb6","value":0.252},{"configuration":"Ling 3.1 Flash","model_id":"m-f74476b1532b3a4f","normalized_unit_value":"1/10","observation_id":"d7e50c008783e8a4b24ebddf0211490f","value":0.1},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"37/250","observation_id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","value":0.148},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"57/500","observation_id":"c287cda1f48a055602b23132b6d39858","value":0.114},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"1/20","observation_id":"d7e2c6c9bb5b766f6065fc5409408479","value":0.05},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"39/250","observation_id":"93e5cef565694d0b6ece46832b2fc2e8","value":0.156}],"protocol_id":"82ccf77be7c53102e3bec8a8","unit":"比例"},"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481":{"benchmark":"Chartography","benchmark_version":"current published task suite","members":[{"configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"71","observation_id":"600568174d3d7470e372c6b4483b5633","value":71.0},{"configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"358/5","observation_id":"13ad1c51a1b24d02661158660b44e5e0","value":71.6},{"configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","model_id":"m-98ee7005cf45b532","normalized_unit_value":"231/5","observation_id":"cee957e5db92decea0ed1ccab6ac32b5","value":46.2},{"configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","model_id":"m-cfe0652e9d685783","normalized_unit_value":"663/10","observation_id":"6d4e060e399a72c43217ec5b1d6a09b6","value":66.3},{"configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"145/2","observation_id":"98879b4fa09dc7b005731381bea066e0","value":72.5}],"protocol_id":"bc37dbdbfe633db52e1c0b8e","unit":"%"},"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f":{"benchmark":"Chartography","benchmark_version":"current published task suite","members":[{"configuration":"Sonnet 5","model_id":"m-2cd37409a483435a","normalized_unit_value":"78/5","observation_id":"a4e7fdc6e1f2295b627a11265f28666d","value":15.6},{"configuration":"GPT-6 Sol","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"268/5","observation_id":"961e2599d1347e8ad57a29581acdb50c","value":53.6},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"417/5","observation_id":"1ce2cb7b64555f9b41b3b5999f8e70ba","value":83.4},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"442/5","observation_id":"a7fcafd577ada1f13084bdc5d25fc8ab","value":88.4},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"322/5","observation_id":"d2e4430a89fa4e3972bc81f62bbdad07","value":64.4},{"configuration":"Sonnet 5.5","model_id":"m-d123610543e5f5ed","normalized_unit_value":"308/5","observation_id":"91fafd2de745057ad9bf6ab81bdb8068","value":61.6}],"protocol_id":"363770683a232ce7f97c2d56","unit":"%"},"489ceb2a92a2c6667e26bfb51444560c46e48f85c7cadff422bc3bc2323afee5":{"benchmark":"Terminal-Bench","benchmark_version":"4.0","members":[{"configuration":"Haiku 5.5","model_id":"m-016da3c7828e329e","normalized_unit_value":"196/5","observation_id":"be88a9ace82ed4a1faf1b6eaa2b8a0bf","value":39.2},{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"373/10","observation_id":"5ba80265411deda6cb00777e57f78095","value":37.3},{"configuration":"Sonnet 5","model_id":"m-2cd37409a483435a","normalized_unit_value":"103/10","observation_id":"8ab92ff9c703946a72390a40d74174e5","value":10.3},{"configuration":"GPT-6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"579/10","observation_id":"83392963b6824da10354750a678d58cf","value":57.9},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"523/10","observation_id":"986d51b676eb20ae4c487a9e43e01df2","value":52.3},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"42","observation_id":"3150cf0f12ca60dc3d5a4b9a023f31b8","value":42.0},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"279/5","observation_id":"03754527ade1d4e46c831eb614495a2b","value":55.8},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"332/5","observation_id":"0e04d5cca5cf4605416149e9e081590e","value":66.4},{"configuration":"Sonnet 5.5 For reference","model_id":"m-d123610543e5f5ed","normalized_unit_value":"353/5","observation_id":"2b76f8801c1b2fa6370b3cd152e2d18c","value":70.6},{"configuration":"GPT-6 Luna","model_id":"m-e5d076977e951607","normalized_unit_value":"82/5","observation_id":"c6b0cf71a2d3dc34173c226f0e2d751d","value":16.4}],"protocol_id":"25f728ad48355959443be093","unit":"%"},"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805":{"benchmark":"MCP Atlas","benchmark_version":"1000 tasks","members":[{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"366/5","observation_id":"a05fb15039bcbd3145b6e77f10451f52","value":73.2},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"382/5","observation_id":"f29e77fd200946b25fe37a0df78120da","value":76.4}],"protocol_id":"a64252d4a51c4ff3960d90c0","unit":"%"},"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9":{"benchmark":"tau3-Banking","benchmark_version":"1.0.1","members":[{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"111340206185567/2000000000000000","observation_id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","value":0.0556701030927835},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"23840206185567/312500000000000","observation_id":"a1e714634bbaa4b289e4726c4bb2c9e9","value":0.0762886597938144},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"179381443298969/1000000000000000","observation_id":"3cd21f0f8001e029d7b19208e29ba099","value":0.179381443298969},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"97938144329897/500000000000000","observation_id":"7b39b47195e57aecded4286410a368f3","value":0.195876288659794},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"639175257731959/10000000000000000","observation_id":"a7dc3e52deea9234fc190c75c4007564","value":0.0639175257731959},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"187628865979381/1000000000000000","observation_id":"b9bbbaa648034918f0dabaffc06da3bd","value":0.187628865979381},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"29381443298969/250000000000000","observation_id":"c4bd6d87d676b473fe7bbadf8b7238e4","value":0.117525773195876},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"52061855670103/250000000000000","observation_id":"dc8c6983f496e505270c3f1d2e08a1c1","value":0.208247422680412},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"224742268041237/500000000000000","observation_id":"ceb2e04d7badea9a2869bdbce0586635","value":0.449484536082474},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"207216494845361/500000000000000","observation_id":"388b55b5431573f02c7a1415569151aa","value":0.414432989690722},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"75257731958763/500000000000000","observation_id":"d0f6a7a8e824abf26de5a2c3eb1a8991","value":0.150515463917526},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"202061855670103/1000000000000000","observation_id":"1b7ece54bb0355e69302fd5373585994","value":0.202061855670103},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"82061855670103/200000000000000","observation_id":"b1886f41be6be37742b893448f0c7326","value":0.410309278350515},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"65979381443299/500000000000000","observation_id":"69ddb67fade6ab2b7cbae1dad6474f61","value":0.131958762886598},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"52061855670103/250000000000000","observation_id":"e330721d27c5f1f48579252fe1951f5e","value":0.208247422680412},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"28659793814433/100000000000000","observation_id":"9ed3eb345da97e474521a1cd387dd7bc","value":0.28659793814433},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"175257731958763/1000000000000000","observation_id":"9202014a7bc9d24df89fe2534ffc35bc","value":0.175257731958763},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"21520618556701/62500000000000","observation_id":"a5800dd77226525d5b0005d3b9d5c59c","value":0.344329896907216},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"232989690721649/1000000000000000","observation_id":"87111691a6a525e89691d7ee783ff6e3","value":0.232989690721649},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"97938144329897/2500000000000000","observation_id":"09489dd17969a18613dce0aff6c534ba","value":0.0391752577319588},{"configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","model_id":"m-7b2a8fe101678ebd","normalized_unit_value":"111340206185567/2000000000000000","observation_id":"24158d9869238dabbd834ff7fd267dc4","value":0.0556701030927835},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"239175257731959/500000000000000","observation_id":"1cd94c3f93e80676324cc2282ca34aa4","value":0.478350515463918},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"597938144329897/10000000000000000","observation_id":"d4ac6d97f1d2caee1aac75fe86cb1d41","value":0.0597938144329897},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"247422680412371/5000000000000000","observation_id":"57be0e2f641053db7912c66d73e8ba7b","value":0.0494845360824742},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"9020618556701/62500000000000","observation_id":"767893d84c1938b3d363a56f8204eb3b","value":0.144329896907216},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"192783505154639/500000000000000","observation_id":"c993b26e42ea6045232f1d9e621ef2fa","value":0.385567010309278},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"59020618556701/125000000000000","observation_id":"a5ec26f61468360fa82a5667cce554db","value":0.472164948453608},{"configuration":"Motif 3","model_id":"m-9e1e0f605d47cf02","normalized_unit_value":"352577319587629/1000000000000000","observation_id":"479061fd58fe09f756e592ab99fa1642","value":0.352577319587629},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"88659793814433/1000000000000000","observation_id":"612f78abda8cf62d29f104d80c127636","value":0.088659793814433},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"319587628865979/1000000000000000","observation_id":"7e3c134e84deebeefe2308a65aef1578","value":0.319587628865979},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"142268041237113/1000000000000000","observation_id":"5c824422e02d6ae8e8e0e7805866dd38","value":0.142268041237113},{"configuration":"G9v3-39A5B","model_id":"m-af0edba2f76c80d6","normalized_unit_value":"220618556701031/1000000000000000","observation_id":"76481bf229eeeddb8b44a15648197449","value":0.220618556701031},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"36340206185567/125000000000000","observation_id":"22b4395fa3e0e77e06e9fcfe630f8789","value":0.290721649484536},{"configuration":"Solar Open2 250B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"8659793814433/40000000000000","observation_id":"e24b9f9c121b7281ff7f37b4b404f354","value":0.216494845360825},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"235051546391753/1000000000000000","observation_id":"4a499df1f84b1eaf0a172cfe97a95f4c","value":0.235051546391753},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"175257731958763/1000000000000000","observation_id":"0456438c8b02eed2e908e4593e41dcf0","value":0.175257731958763},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"115463917525773/1000000000000000","observation_id":"77cc539d6e52970112947681c17a7cf5","value":0.115463917525773},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"138144329896907/500000000000000","observation_id":"33c50c79dd62f0151f2db3df92a057bc","value":0.276288659793814},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"100618556701031/200000000000000","observation_id":"0034bad285685afff6ae0af098ac3976","value":0.503092783505155},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"480412371134021/1000000000000000","observation_id":"02b14b2382923c4835faa17152c2d94b","value":0.480412371134021},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"59020618556701/125000000000000","observation_id":"9a44264bd8b89c341b6a8bd9496be3bf","value":0.472164948453608},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"251546391752577/1000000000000000","observation_id":"84d4042d1c82bfee0ac491e53b4fa917","value":0.251546391752577},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"360824742268041/5000000000000000","observation_id":"ad5761dce64ad5d75bd8fe071f2c9120","value":0.0721649484536082},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"228865979381443/1000000000000000","observation_id":"a00dadd2de4cef1d6ffa67d4028686b0","value":0.228865979381443},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"342268041237113/1000000000000000","observation_id":"348ef206dd0fa625708f257a200455f5","value":0.342268041237113},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"160824742268041/1000000000000000","observation_id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","value":0.160824742268041},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"152577319587629/1000000000000000","observation_id":"0494d5d3e701ac0729bd55524064b2fa","value":0.152577319587629},{"configuration":"Agnes 2.5 Pro Beta","model_id":"m-e060cf24e4176269","normalized_unit_value":"71340206185567/200000000000000","observation_id":"569f96b9f64d6f89c3e721fb138a9646","value":0.356701030927835},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"11340206185567/25000000000000","observation_id":"24d253e4fe3d6de3cc614f0a892166b1","value":0.45360824742268},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"175257731958763/1000000000000000","observation_id":"62cc0b9af0e4241c44026a470cdd621c","value":0.175257731958763},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"202061855670103/500000000000000","observation_id":"66284fe19cdf80bb97a995a814e23753","value":0.404123711340206},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"197938144329897/500000000000000","observation_id":"20bfc5477d34134d5c7c99539fd481a0","value":0.395876288659794},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"61340206185567/125000000000000","observation_id":"960dd18162897dbffed3123d8b9728ee","value":0.490721649484536}],"protocol_id":"416e531f10637380f0b1b24d","unit":"比例"},"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493":{"benchmark":"MultiChallenge","benchmark_version":"current published task suite","members":[{"configuration":"Claude Haiku 5.5","model_id":"m-016da3c7828e329e","normalized_unit_value":"6767/100","observation_id":"5d9b045cf9ce01ea061077e1b2e84c1e","value":67.67},{"configuration":"gpt 6.1 Sol","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"1639/20","observation_id":"2a39ee205e56064a7ebeddb7db41c3ee","value":81.95},{"configuration":"GPT 6.1 Terra","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"371/5","observation_id":"c65e268d91030372566c3ec5c7a1ab64","value":74.2},{"configuration":"Gemini 4 Argon","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"787/10","observation_id":"2db7888810c9c42969f9152c713f9d6e","value":78.7},{"configuration":"Muse Spark Open","model_id":"m-db208dec31031420","normalized_unit_value":"371/5","observation_id":"fc1a81b2311badda3a38cf9a6a28b5f4","value":74.2}],"protocol_id":"668b5eec1bd41046308d40fb","unit":"%"},"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec":{"benchmark":"MMMU-Pro","benchmark_version":"Standard 10","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"83","observation_id":"40f8d3f23ce7ba4f99898294659f1d24","value":83.0},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"392/5","observation_id":"b27c843f9cde1b388c12022729f0ceb0","value":78.4},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"807/10","observation_id":"707e27b848a443e1f331abbd2d46a87f","value":80.7}],"protocol_id":"c9f689bb3d88b096fb77bc89","unit":"%"},"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"GPT 5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"73","observation_id":"975826e2b4130a8d077f072f1973ca40","value":73.0},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"70","observation_id":"985255818b7c5bb489a1bbeb1bc94859","value":70.0},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"283/5","observation_id":"9e8a69deb5a1faa9380e095362bd2e2c","value":56.6},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"108/5","observation_id":"2206795ef677e458e6f834638793b28d","value":21.6},{"configuration":"Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"59","observation_id":"ec4b61209e562c5ed86fc0f393950588","value":59.0}],"protocol_id":"4c5c14ed1147651f8f95c7e1","unit":"%"},"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862":{"benchmark":"GDPval-AA","benchmark_version":"2.1","members":[{"configuration":"Haiku 5.5","model_id":"m-016da3c7828e329e","normalized_unit_value":"1620","observation_id":"e5de55a774fc4f793b69c43a1c3f15d2","value":1620.0},{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"1588","observation_id":"8aeb7f071b6b514a7531446104ba33a1","value":1588.0},{"configuration":"Sonnet 5","model_id":"m-2cd37409a483435a","normalized_unit_value":"1449","observation_id":"fcb5b9889977ea6f8824ae1c412b655a","value":1449.0},{"configuration":"GPT-6 Sol","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"1487","observation_id":"8b095392eb68858c1d89a63205401933","value":1487.0},{"configuration":"GPT-6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"1542","observation_id":"c006e5f0249adc5c225cc895f365ee07","value":1542.0},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"1708","observation_id":"3e2aa560fb85d043c269e4880c8d1ed7","value":1708.0},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"1735","observation_id":"c754242c94abd65cff43bbb52150d354","value":1735.0},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"1846","observation_id":"2674db64c27edbe09263f61e9c5887af","value":1846.0},{"configuration":"Sonnet 5.5 For reference","model_id":"m-d123610543e5f5ed","normalized_unit_value":"1840","observation_id":"4ea887aa7cc58c78a364e0a2b65d69bf","value":1840.0},{"configuration":"GPT-6 Luna","model_id":"m-e5d076977e951607","normalized_unit_value":"1437","observation_id":"c37eddc526d0ea268561864cec601165","value":1437.0}],"protocol_id":"34c76587afd7902bbcf38773","unit":"分"},"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1":{"benchmark":"IFBench","benchmark_version":"current published task suite","members":[{"configuration":"DeepSeek-V4-Flash-0731","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"396/5","observation_id":"bb7586265eb9f4ee440c1de9eb72708e","value":79.2},{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"791/10","observation_id":"e1daf1b784cda878658887e206aebf2f","value":79.1},{"configuration":"Qwen3.8-27B","model_id":"m-c947385433db3f03","normalized_unit_value":"159/2","observation_id":"860832a18411b625ebef7842433113ca","value":79.5},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"813/10","observation_id":"1bcc725977832eb3960634353511e9be","value":81.3}],"protocol_id":"0711b41a3607466304f72d7f","unit":"%"},"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329":{"benchmark":"LiveCodeBench","benchmark_version":"release_v6","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"589/10","observation_id":"2f001b0be2d50fef9306d2b3f9f1569b","value":58.9},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"199/5","observation_id":"445a31daff84293600dd4a2900cbf97d","value":39.8},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"691/10","observation_id":"97029c45627dc4bbce9101042fef65e2","value":69.1},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"421/10","observation_id":"ff54ed0b44635ad5d8a708f3d17d4063","value":42.1}],"protocol_id":"9b21b259ed1433cdf48c1913","unit":"%"},"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917":{"benchmark":"APEX-Agents","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6 Sol / official Hy4 comparison, APEX-Agents pass@1","model_id":"m-0510680735d3c79a","normalized_unit_value":"379/10","observation_id":"85b0ca5b600459af0704f6882d2f67f7","value":37.9},{"configuration":"Hy4 preview / official APEX-Agents pass@1","model_id":"m-383f40728e3e5259","normalized_unit_value":"371/10","observation_id":"7155f49a20e936c90be455db70c3df42","value":37.1},{"configuration":"Claude Opus 5 / official Hy4 comparison, APEX-Agents pass@1","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"209/5","observation_id":"f541c2fdbde72cb43c74ebe8a432a7bf","value":41.8},{"configuration":"Qwen3.8 Max (0902) / official Hy4 comparison, APEX-Agents pass@1","model_id":"m-825009614c016c1d","normalized_unit_value":"34","observation_id":"ec578f1a940281fdcb27687071dac7ac","value":34.0},{"configuration":"Kimi K3 / official Hy4 comparison, APEX-Agents pass@1","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"186/5","observation_id":"7b257a4e69fd26a5efca3f99844cb10c","value":37.2},{"configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","model_id":"m-c65e3e17492500e9","normalized_unit_value":"381/10","observation_id":"21a8aa59fcb39de6e961594401b8acf8","value":38.1},{"configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","model_id":"m-f936349cbdb1def9","normalized_unit_value":"162/5","observation_id":"76dde74584cddc5a2d8618786061726c","value":32.4}],"protocol_id":"9421676f6c30b2c29c7fe852","unit":"%"},"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6":{"benchmark":"Terminal-Bench-Science","benchmark_version":"0.1.0","members":[{"configuration":"Claude Haiku 5.5 / Mini-SWE-agent","model_id":"m-016da3c7828e329e","normalized_unit_value":"10","observation_id":"0e248a56228a0b9ac7b762f2f3b00fc9","value":10.0},{"configuration":"GPT-5.6 Sol / Mini-SWE-agent","model_id":"m-0510680735d3c79a","normalized_unit_value":"20","observation_id":"5555c117d96b482e6d1bb0b304cbbefd","value":20.0},{"configuration":"GPT-6.1 Sol / Mini-SWE-agent","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"2643/50","observation_id":"77175d98c0074eb53a919fa599931946","value":52.86},{"configuration":"Gemini 3.7 Flash / Mini-SWE-agent","model_id":"m-16a8dce2a6e7167a","normalized_unit_value":"357/50","observation_id":"174d2abbe0e35fdfac2a047ae2fa2208","value":7.14},{"configuration":"Muse Wonder Full / Mini-SWE-agent","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"1857/50","observation_id":"dbbbd82ee29a731632cb4d28c2efdeea","value":37.14},{"configuration":"Claude Sonnet 5 / Mini-SWE-agent","model_id":"m-2cd37409a483435a","normalized_unit_value":"571/100","observation_id":"01d54ae905ebb77f621f6cbf30eb0930","value":5.71},{"configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"429/100","observation_id":"5a0b6bca7e87cdf2b14ffe16a27cf187","value":4.29},{"configuration":"GPT-6 Sol / Mini-SWE-agent","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"30","observation_id":"f1163968f642895b63614f184b419fdc","value":30.0},{"configuration":"Grok 4.6 / Mini-SWE-agent","model_id":"m-3c17c0d3ca7f1abe","normalized_unit_value":"1143/100","observation_id":"545c704b3f9b5b5c3a00fbe6211a666c","value":11.43},{"configuration":"Gemini 3.8 Flash / Mini-SWE-agent","model_id":"m-422e5ba87859dd55","normalized_unit_value":"857/100","observation_id":"4c2b94e36c85b9d8f2e3d83490a34034","value":8.57},{"configuration":"GPT-6 Astra / Mini-SWE-agent","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"3143/50","observation_id":"33b37c700e39b0d0c6a88f59bf14c960","value":62.86},{"configuration":"Claude Opus 5 / Mini-SWE-agent","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"1357/50","observation_id":"12c1a7b1c4eb18f0409ef269987fa5ca","value":27.14},{"configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","model_id":"m-60cd64beda366768","normalized_unit_value":"429/100","observation_id":"f38424924d0f84fbc9a6a159bb2bb1d4","value":4.29},{"configuration":"GPT-5.6 Luna / Mini-SWE-agent","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"0","observation_id":"85fbb43a0855c52959cd8ec4413297f0","value":0.0},{"configuration":"Kimi K3.1 Pro / Mini-SWE-agent","model_id":"m-66c1aae520b7679a","normalized_unit_value":"3143/100","observation_id":"58dd7cae94574219da61327aad86af6e","value":31.43},{"configuration":"Muse Wonder / Mini-SWE-agent","model_id":"m-68480fc0423c5034","normalized_unit_value":"1143/50","observation_id":"adff8e814a52d75cbf97b5acfa5177a4","value":22.86},{"configuration":"Grok 4.7 / Mini-SWE-agent","model_id":"m-7187cf2b01d41935","normalized_unit_value":"10","observation_id":"5f766555b014c71ff2d651ccb2fc76d7","value":10.0},{"configuration":"Gemini 3.6 Flash / Mini-SWE-agent","model_id":"m-77b5dbc317075e9d","normalized_unit_value":"429/100","observation_id":"1632686674422c3a2065b7adb4e89430","value":4.29},{"configuration":"Gemini 4 Argon / Mini-SWE-agent","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"4429/100","observation_id":"50ddd51f0edebe68ad08c74703710c38","value":44.29},{"configuration":"Claude Fable 5 / Mini-SWE-agent","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"1571/100","observation_id":"f5cc9d3986b503f55588d97bf5fde40a","value":15.71},{"configuration":"Gemini 3.5 Flash / Mini-SWE-agent","model_id":"m-8d99798a7e24690c","normalized_unit_value":"571/100","observation_id":"72ae1d1de2eb58dbc8e135c9056f08ba","value":5.71},{"configuration":"Claude Fable 5.1 / Mini-SWE-agent","model_id":"m-98ee7005cf45b532","normalized_unit_value":"40","observation_id":"2012b47ec868b52e455d3d8d66a1a48a","value":40.0},{"configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"0","observation_id":"90672f02fa6c8adab2a2801b7c2aedb3","value":0.0},{"configuration":"Qwen3.8 Max / Mini-SWE-agent","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"643/50","observation_id":"a18d9b692129832a6701c8778a0404f3","value":12.86},{"configuration":"Kimi K3 / Mini-SWE-agent","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"143/100","observation_id":"f2b3352d05d0abb45473dcad237d2a14","value":1.43},{"configuration":"Beam 1.0 / Mini-SWE-agent","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"143/100","observation_id":"04366109cc25918e59cfe4140983a31b","value":1.43},{"configuration":"Mercury 2.5 / Mini-SWE-agent","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"0","observation_id":"80525908530cba67da8858ed258e1742","value":0.0},{"configuration":"Muse Spark 1.3 / Mini-SWE-agent","model_id":"m-c0580b01221ed260","normalized_unit_value":"429/100","observation_id":"6e4fcbf667751f178fabfb4106350208","value":4.29},{"configuration":"GLM-5.3 / Mini-SWE-agent","model_id":"m-c65e3e17492500e9","normalized_unit_value":"571/100","observation_id":"6cbde617319ab4da7917bf56251c3a2e","value":5.71},{"configuration":"Qwen3.8 27B / Mini-SWE-agent","model_id":"m-c947385433db3f03","normalized_unit_value":"0","observation_id":"d437bf4a69b6a24c59bada231457a5bc","value":0.0},{"configuration":"Claude Opus 4.8 / Mini-SWE-agent","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"429/100","observation_id":"540adb8175ab28d08c531a963e348646","value":4.29},{"configuration":"GLM-5.3-Flash / Mini-SWE-agent","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"143/50","observation_id":"4e220e5978b3977fd57d2c06a76a9766","value":2.86},{"configuration":"Claude Opus 5.5 / Mini-SWE-agent","model_id":"m-cfe0652e9d685783","normalized_unit_value":"2357/50","observation_id":"21af8993ff5f3e38ecc3caf66271b208","value":47.14},{"configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","model_id":"m-d123610543e5f5ed","normalized_unit_value":"2357/50","observation_id":"0fd84fb90c507070cc93b2671c1dacd9","value":47.14},{"configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"143/50","observation_id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","value":2.86},{"configuration":"GPT-6 Luna / Mini-SWE-agent","model_id":"m-e5d076977e951607","normalized_unit_value":"429/100","observation_id":"a67f5a0d13a3ef07b45639befa77b60c","value":4.29},{"configuration":"Step 5 Preview / Mini-SWE-agent","model_id":"m-f892623396612d9b","normalized_unit_value":"143/50","observation_id":"f80ea7a9c8b1ad0f866390a070dec6ae","value":2.86},{"configuration":"GPT-5.6 Terra / Mini-SWE-agent","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"10","observation_id":"9d8fe865edc8da70092f0b94926485a3","value":10.0},{"configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","model_id":"m-f936349cbdb1def9","normalized_unit_value":"429/100","observation_id":"0898509511ccb710480bf98d3742bbb0","value":4.29}],"protocol_id":"dbbbcaf695179b55af0fc57c","unit":"%"},"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"297/10","observation_id":"d2394a533761e9936e4a49c1dfe512c8","value":29.7},{"configuration":"Fable5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"291/10","observation_id":"e6769544adbda039071845c194ca792d","value":29.1},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"273/10","observation_id":"fb72cb8ea716369f0491dda0a2c1f8cc","value":27.3},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"71/5","observation_id":"4ea5a3b77e03d03f1422a6728fc0742b","value":14.2},{"configuration":"Opus4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"136/5","observation_id":"d58fa04ae1f8f5897c36ae81b8ac3dcf","value":27.2}],"protocol_id":"8034c6f82ee403f4844044bd","unit":"%"},"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919":{"benchmark":"Terminal-Bench","benchmark_version":"4.0","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"82070707070707/250000000000000","observation_id":"7706b3202bdd1e09b6c5e4786d4bb3e4","value":0.328282828282828},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"560606060606061/1000000000000000","observation_id":"d04f5531bb77f209689af5059ec9471d","value":0.560606060606061},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"0","observation_id":"c56ed05cd995d4a9a967c8a8edd7c23f","value":0},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"0","observation_id":"f435dfe97ff078f0f04715c2a33b7561","value":0},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"138888888888889/250000000000000","observation_id":"35604c8a1315056deeba470a2491f933","value":0.555555555555556},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"570707070707071/1000000000000000","observation_id":"fc581128088f8e6a2ca283411c75c69c","value":0.570707070707071},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"7e0782bec5efa3fc907fbba656122a79","value":0.00505050505050505},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"0","observation_id":"585ae90ea798ef43916073acc85d178e","value":0},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","value":0.00505050505050505},{"configuration":"JT-4.1 Flash 236B A21B (Reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"303030303030303/10000000000000000","observation_id":"8ad3da1ee13871ee0748612b7b0f2202","value":0.0303030303030303},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"fbda3169b181acd989a4524a5ca42c7f","value":0.0101010101010101},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"0","observation_id":"01e4bcd3555fdf1970b9b729ff28db44","value":0},{"configuration":"Nemotron 3.5 Bolt 74B-A9B","model_id":"m-3ad11f2b6253b364","normalized_unit_value":"0","observation_id":"614dafc454b6d64e61b2f25573ec4717","value":0},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"63131313131313/500000000000000","observation_id":"4473f9853fd2b2892365909edf9631fb","value":0.126262626262626},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"0","observation_id":"49929d3422a54a6833b2d573268b4f8f","value":0},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"196969696969697/1000000000000000","observation_id":"477576b238b66114c394a26ca4a20345","value":0.196969696969697},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"590909090909091/1000000000000000","observation_id":"cf87d9afdb957adc6bd1bb79bf9aa131","value":0.590909090909091},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"0","observation_id":"47e9d9d6a84d129959dab9eed70b71bc","value":0},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"3126ca26ac0a6c7dab9a15033082575d","value":0.0101010101010101},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"227272727272727/1000000000000000","observation_id":"8ed311eb703cfe1b731155e091addc7e","value":0.227272727272727},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"121212121212121/1000000000000000","observation_id":"cdfb425c1664f272548a144cf79359c8","value":0.121212121212121},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"167929292929293/250000000000000","observation_id":"1c21aa90c01783fcd785c617ba4d90f9","value":0.671717171717172},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"26010101010101/50000000000000","observation_id":"6fe9193a0e51ae328fa8f19ad476b5c1","value":0.52020202020202},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"0","observation_id":"27a6f0107d1b80cc1605307626b6cc35","value":0},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"0","observation_id":"9fef6d5b9c29a36ef4ec13143b2d54ae","value":0},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"84fdb274483242a5de60a1387cb766c1","value":0.0101010101010101},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"128787878787879/500000000000000","observation_id":"a073820ddc97f09d7966e361779cf68e","value":0.257575757575758},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"a84f629da93a24a910222138960c0c05","value":0.0101010101010101},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"0","observation_id":"b740ee2b3c6f41e732abe95d92f63313","value":0},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"247474747474747/1000000000000000","observation_id":"038de3c11daf5dbf02e5d1c9dd11ccd7","value":0.247474747474747},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"0","observation_id":"e5e132be2e915d3fe97356bbd06869f9","value":0},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"0","observation_id":"732ce0d4c4b93e00640afbdf97e81cd3","value":0},{"configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","model_id":"m-7b2a8fe101678ebd","normalized_unit_value":"0","observation_id":"2b6eac192b734b825e56a7c642b341b7","value":0},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"570707070707071/1000000000000000","observation_id":"58499c853c2644c7eb9a4bc16fc208c6","value":0.570707070707071},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"388888888888889/1000000000000000","observation_id":"901770e3927f8efbffa8b7b6dc69571d","value":0.388888888888889},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"148989898989899/500000000000000","observation_id":"37a1dedbc40bda682abb927c484863ce","value":0.297979797979798},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"ed0927341bbf29454c5f15ec4d2315cd","value":0.00505050505050505},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"128787878787879/500000000000000","observation_id":"3a8fd5af00baca1a19eb582e1a8c4ab5","value":0.257575757575758},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"0","observation_id":"7ed61dbea4d39e1cec33826bfcfe4762","value":0},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"0","observation_id":"027860a90cc2579f5f5af8c36e50c87e","value":0},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"0","observation_id":"d43f0f17f32dc60cbed0bdb79fe20eb1","value":0},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"26010101010101/50000000000000","observation_id":"71f828c6aa1f518ef08599ca0a1b0570","value":0.52020202020202},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"803b4920eccad03514634ebd3575a2d8","value":0.00505050505050505},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"82070707070707/250000000000000","observation_id":"cf0c6cf90a55894b4c003eacdea2ccb7","value":0.328282828282828},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"0","observation_id":"cd7e16f694d6c0543ca83b3b5dd9358b","value":0},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"22a0f772949251594cd3a25e02ef9752","value":0.00505050505050505},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"1344e9967afa11034830d610f592ed83","value":0.0101010101010101},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"303030303030303/10000000000000000","observation_id":"7a6b6215e85870e72209d73fd4101db4","value":0.0303030303030303},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"186868686868687/500000000000000","observation_id":"b9ebef353db3bd587384633039b7a5bb","value":0.373737373737374},{"configuration":"Mercury 2.5","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"0","observation_id":"179a39191c3950939aafcb199439b020","value":0},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"af8dab2781940ab31c2449f02ac9e502","value":0.00505050505050505},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"0","observation_id":"017b50b4eb6f553388684608c12adbeb","value":0},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"5a757d2c757812d80a306fe7007417cd","value":0.0101010101010101},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"419191919191919/1000000000000000","observation_id":"eaa12d2968deb06d22cebe6ac9936af8","value":0.419191919191919},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"138888888888889/2500000000000000","observation_id":"bb15f242cbb67b0a52ecb1efaeaf802e","value":0.0555555555555556},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"169191919191919/500000000000000","observation_id":"e62fa5ce7d9f527df294b117e42de68a","value":0.338383838383838},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"82070707070707/250000000000000","observation_id":"eca9bd0f6ebf44ed64158c0a214f9271","value":0.328282828282828},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"148989898989899/250000000000000","observation_id":"322131c4aa07af285df46f472061072d","value":0.595959595959596},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"0","observation_id":"cafc26cd33f16c1828c33bad26ea21a7","value":0},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"159090909090909/250000000000000","observation_id":"620eabc34583900c0a5610d2452e8c0e","value":0.636363636363636},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"101010101010101/20000000000000000","observation_id":"87cb9fb97ee59be625938b752ce0bb36","value":0.00505050505050505},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"9469696969697/625000000000000","observation_id":"612528520707bc7ebd5f35810f7efeb6","value":0.0151515151515152},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"0","observation_id":"46a72c291abf86482bd30a1de9ec49ac","value":0},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"101010101010101/5000000000000000","observation_id":"ab24b7c89863efc898bb5b0dc33b7ea8","value":0.0202020202020202},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"348484848484849/1000000000000000","observation_id":"a7be2038bfeb07e68eff4c9e5fd86646","value":0.348484848484849},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"101010101010101/1250000000000000","observation_id":"6d1fe83c673cc3e53108a1f798f9857e","value":0.0808080808080808},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"651515151515151/1000000000000000","observation_id":"8dde2603c050365ba8458b5b95c47df3","value":0.651515151515151},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"252525252525253/1000000000000000","observation_id":"0978fbc73a68e2ee70d02c130e0d1cff","value":0.252525252525253},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"63131313131313/500000000000000","observation_id":"f54a9b9328eeb87160c6fecff9cdb2e5","value":0.126262626262626},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"54cd2ea0dc02b5047f97de0db07066fb","value":0.0101010101010101},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"131313131313131/1000000000000000","observation_id":"b334bad67fbd4a96eed8e669e0111fbb","value":0.131313131313131},{"configuration":"Ling 3.1 Flash","model_id":"m-f74476b1532b3a4f","normalized_unit_value":"333333333333333/1000000000000000","observation_id":"070c8b1a9e5fb50b3ddeb6aba50faf78","value":0.333333333333333},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"333333333333333/1000000000000000","observation_id":"c3c1a9e38c3fc762089faf1cd02d09ab","value":0.333333333333333},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"141414141414141/1000000000000000","observation_id":"4f209459e33bb2ec173abeb800f71be3","value":0.141414141414141},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"101010101010101/10000000000000000","observation_id":"7620c0fb57936004d0ff80edb6675f9f","value":0.0101010101010101},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"111111111111111/1000000000000000","observation_id":"8981298b68f7a13ad078d01ac54f2b30","value":0.111111111111111}],"protocol_id":"daf4f954f79ad943ce53905c","unit":"比例"},"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e":{"benchmark":"MCP Atlas","benchmark_version":"1000 tasks","members":[{"configuration":"gpt-5.6 (sol)","model_id":"m-0510680735d3c79a","normalized_unit_value":"409/5","observation_id":"938965e9ee08fb381a111173123203b2","value":81.8},{"configuration":"Inkling-small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"396/5","observation_id":"72495a708e9f00af5b2d17370ac07628","value":79.2},{"configuration":"claude-opus-5 (xhigh)","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"429/5","observation_id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","value":85.8},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"833/10","observation_id":"37801ada9ba8d94f8b7e8f5a6667ee11","value":83.3},{"configuration":"Gemini 3.5 Flash (high)","model_id":"m-8d99798a7e24690c","normalized_unit_value":"418/5","observation_id":"44434b3879169ad5c54a4316dc8731a9","value":83.6},{"configuration":"kimi-k3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"823/10","observation_id":"b1436c94b33fa58416664a5d31b199a9","value":82.3},{"configuration":"Inkling (xHigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"76","observation_id":"73ec2c70e1808cc25e2b747786aa5d99","value":76},{"configuration":"GLM 5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"421/5","observation_id":"442d36ffd8ae4e2ebc285d09f51efdf2","value":84.2},{"configuration":"claude-opus-4-8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"411/5","observation_id":"4cec2c81b7ca1a60e4ad36412d5896a7","value":82.2},{"configuration":"Qwen3.8-2.4T-A95B (xHigh)","model_id":"m-fb920161af1862b7","normalized_unit_value":"169/2","observation_id":"7b2615183b2725bb4693774f30af4a5b","value":84.5}],"protocol_id":"89dd905180d24f13af9f3f8e","unit":"%"},"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03":{"benchmark":"AA-AnalystAgent","benchmark_version":"current AA pass^5","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"43/80","observation_id":"a7b164121b0a9a4769bf158d9212d770","value":0.5375},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"1/2","observation_id":"9b9634fd2b2367e076a491de1e4f02e6","value":0.5},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"11/40","observation_id":"6890e628f5988bd5a798a680de5a69d4","value":0.275},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"37/80","observation_id":"9cd6a63a2e9021c4431ddd906f3978b5","value":0.4625},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"41/80","observation_id":"15975e3d42ab5d82a67bcbe277dd1b83","value":0.5125},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"1/8","observation_id":"ef6364df2840f04f7c6f211bd995980d","value":0.125},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"31/80","observation_id":"8b93841abeaaa5b29864bb557e6a9da7","value":0.3875},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"21/40","observation_id":"1c70059543aa3443966891f5d87280d7","value":0.525},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"39/80","observation_id":"4f1262f25c1f33cd755aebe2e08018dd","value":0.4875},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"9/20","observation_id":"cee5a16789123af6ef334475ebff73a7","value":0.45},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"13/80","observation_id":"c283db79a5fa78b491d2c36ac505fd73","value":0.1625},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"23/40","observation_id":"172f756dc0c16d9248bcf51116651a4b","value":0.575},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"1/16","observation_id":"2275301a736d0fe2b34ce5c990945f54","value":0.0625},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"19/80","observation_id":"753dbaee803264fefa51f263a84b1509","value":0.2375},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"1/5","observation_id":"809f69850cfb92e35fdef7e6fcd0b405","value":0.2},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"31/80","observation_id":"18b8a0177c3660f9aaf285f2b3571603","value":0.3875},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"1/4","observation_id":"a9057fe2d742aeb1aa0666773aa5958d","value":0.25},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"9/16","observation_id":"1a02ab6144866fd1b960f19d8fd21513","value":0.5625},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"23/40","observation_id":"b14e1b2737d16e7bb80a947abb8a320a","value":0.575},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"1/10","observation_id":"8cef68049263078fc3646fc52bd5e75a","value":0.1},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"3/8","observation_id":"59ce77c04181c398c36c852841fb4ecf","value":0.375},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"7/16","observation_id":"ebd2172910fbd66a3c4a6655aa5bdef4","value":0.4375},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"41/80","observation_id":"e23e94bce569e3668ea49a1736f85e07","value":0.5125},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"7/20","observation_id":"1f372ec6dfb68a53b6ae42bbb816dddf","value":0.35}],"protocol_id":"0fdc20dc341754f0aa205de6","unit":"比例"},"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT 5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"297/10","observation_id":"ebef37315791ccc04717ed5ba0766942","value":29.7},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"291/10","observation_id":"e5e10458ba007e3ac9010def1e16a95d","value":29.1},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"273/10","observation_id":"58a53dd6f007e225a58569d30c64eee8","value":27.3},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"71/5","observation_id":"ffb82b85f11b529dafa9689c16b28131","value":14.2},{"configuration":"Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"136/5","observation_id":"d58257a11e2e15bf0bb878bb097f91ba","value":27.2}],"protocol_id":"2a48ec513964ac2d123d226b","unit":"%"},"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927":{"benchmark":"Vals Code Migration","benchmark_version":"1.1","members":[{"configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","model_id":"m-016da3c7828e329e","normalized_unit_value":"1309/25","observation_id":"c1b5961685ee27b7ccd1b99becf1d743","value":52.36},{"configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"1933/25","observation_id":"e330fa37e52713545b5fec3f612240b8","value":77.32},{"configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"6279/100","observation_id":"3d507b92bee70e08544b9eb65b20267a","value":62.79},{"configuration":"Inkling Small / Published Code Migration 1.1 sandbox","model_id":"m-35c4f287acc76c95","normalized_unit_value":"1043/100","observation_id":"17347953383a652f8a8fabeabf1d07e6","value":10.43},{"configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","model_id":"m-383f40728e3e5259","normalized_unit_value":"1134/25","observation_id":"8b32286773fba058c9bb87cdd6ac0c71","value":45.36},{"configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","model_id":"m-422e5ba87859dd55","normalized_unit_value":"773/20","observation_id":"59c32ec585e3e137a1a5f19c3bda0e98","value":38.65},{"configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"4029/50","observation_id":"bc107d6862df5282c7cf8b968206739b","value":80.58},{"configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","model_id":"m-60cd64beda366768","normalized_unit_value":"791/20","observation_id":"501724f84396d924c42fcaca484ef573","value":39.55},{"configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","model_id":"m-68480fc0423c5034","normalized_unit_value":"1809/25","observation_id":"ab784b677fcf21bc1b248358effdf278","value":72.36},{"configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","model_id":"m-7187cf2b01d41935","normalized_unit_value":"2661/50","observation_id":"575f9b1e7b6ccadadc3e13c6241b21d4","value":53.22},{"configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"354/5","observation_id":"2905dc672ed8db458e1dc8c5a09c04a7","value":70.8},{"configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","model_id":"m-825009614c016c1d","normalized_unit_value":"5719/100","observation_id":"62c5b15d1c0217ab796423502244a099","value":57.19},{"configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"2947/50","observation_id":"3beb1575e4852b44427673ea326ae656","value":58.94},{"configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"2859/100","observation_id":"549f9102d89b3e9a3355e2ef41ef0184","value":28.59},{"configuration":"Inkling / Published Code Migration 1.1 sandbox","model_id":"m-b439c804161fac26","normalized_unit_value":"441/50","observation_id":"2bad0033060ca0f870378d7a7d3d6c5d","value":8.82},{"configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"457/20","observation_id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","value":22.85},{"configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","model_id":"m-c65e3e17492500e9","normalized_unit_value":"5507/100","observation_id":"78dbe912ae9dbb4dfacad05113fef49e","value":55.07},{"configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"1688/25","observation_id":"1080a38736b6ba3768e68ace6a340f30","value":67.52},{"configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"1033/20","observation_id":"efd99741d3524e2d5ab3b679f8da8006","value":51.65},{"configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","model_id":"m-cfe0652e9d685783","normalized_unit_value":"8101/100","observation_id":"d2e15e2796cf44b3aabbae2da4bd5b3f","value":81.01},{"configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","model_id":"m-d123610543e5f5ed","normalized_unit_value":"2172/25","observation_id":"165b56e6a0826f1ed0dc82b37b0885fe","value":86.88},{"configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","model_id":"m-da421acad97d607b","normalized_unit_value":"209/10","observation_id":"fbec8a823687d43757fc0efa8ad26423","value":20.9},{"configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"4879/100","observation_id":"c811b7493c2ca32eedcc762a5f354d58","value":48.79},{"configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","model_id":"m-db208dec31031420","normalized_unit_value":"2689/100","observation_id":"aeba1081b4abe0557cd65f41eca5f8b3","value":26.89},{"configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","model_id":"m-e5d076977e951607","normalized_unit_value":"4667/100","observation_id":"1045763a4cc14f13d71cbc23b34d8b21","value":46.67},{"configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","model_id":"m-f892623396612d9b","normalized_unit_value":"47","observation_id":"68fce01452fe549878e8933d04edaa05","value":47.0},{"configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","model_id":"m-f936349cbdb1def9","normalized_unit_value":"5199/100","observation_id":"7e542189106406c0f4ba769f7753c91e","value":51.99}],"protocol_id":"08d602c28ff5aa288b184edb","unit":"%"},"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1":{"benchmark":"MMMU-Pro","benchmark_version":"Standard 10","members":[{"configuration":"GPT-5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"83","observation_id":"c21818a023d8c8e1946a57f37d71c67d","value":83.0},{"configuration":"Claude Fable 5 (max, w/ fallback)","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"406/5","observation_id":"f6eb2e4c060926f84692b3351b0182f9","value":81.2},{"configuration":"Kimi K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"408/5","observation_id":"9ea024ab5ca2b01fdc71c3f50f184c9c","value":81.6},{"configuration":"Claude Opus 4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"789/10","observation_id":"24b2d438f8c9ef458157a109e7bcb701","value":78.9}],"protocol_id":"6fde3ba0bea9afec4846158a","unit":"%"},"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80":{"benchmark":"MMLU-Pro","benchmark_version":"current maintainer dataset","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"329/5","observation_id":"8519642492aa86c8bdd1ad71a0f84568","value":65.8},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"631/10","observation_id":"0129fdb380d08e0198547b85719a77cd","value":63.1},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"354/5","observation_id":"fdfed7f2ef133f7ec35f427f9ee8f813","value":70.8},{"configuration":"Solar Open 2 250B-A15B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"431/5","observation_id":"27d4cca7113cd0216eb12449ea9867c2","value":86.2},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"326/5","observation_id":"239f2b6a72085e3eb63d599375b5b516","value":65.2}],"protocol_id":"dff05c56b6d8c6ac1cddbc91","unit":"%"},"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958":{"benchmark":"SimpleQA Verified","benchmark_version":"run 1.2.0","members":[{"configuration":"claude-haiku-5-5_max","model_id":"m-016da3c7828e329e","normalized_unit_value":"119/500","observation_id":"a7e4020697adbd00e97487d7c55696e0","value":0.238},{"configuration":"gpt-6.1-sol_max","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"739/1000","observation_id":"509c52385e7c5559b4414a3e4bdb93cc","value":0.739},{"configuration":"gemini-3.7-flash_high","model_id":"m-16a8dce2a6e7167a","normalized_unit_value":"173/250","observation_id":"ab75c5704ac20fa5dc1c93828558d2ce","value":0.692},{"configuration":"gpt-6.1-terra_max","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"573/1000","observation_id":"52a58ef3bf54eb42e218e3c094aec598","value":0.573},{"configuration":"claude-sonnet-5_xhigh","model_id":"m-2cd37409a483435a","normalized_unit_value":"329/1000","observation_id":"0c6f13f860966ddef45cc04368fd5127","value":0.329},{"configuration":"muse-spark-1.2_xhigh","model_id":"m-2f22b07da3af863c","normalized_unit_value":"1507537688442211/2500000000000000","observation_id":"aa61e58d244812473e1af5a8764b4e07","value":0.6030150753768844},{"configuration":"deepseek-v4-flash-0731_max","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"8408408408408409/25000000000000000","observation_id":"f310a7a1d6e944a484a86a2598b55b31","value":0.33633633633633636},{"configuration":"inkling-small_xhigh","model_id":"m-35c4f287acc76c95","normalized_unit_value":"191/1000","observation_id":"62e95547f0f7073efcdd1420b882e4a5","value":0.191},{"configuration":"gpt-6-sol_max","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"607/1000","observation_id":"8a0eac6d7e9eaac51a5eeee920a35f4d","value":0.607},{"configuration":"grok-4.6_high","model_id":"m-3c17c0d3ca7f1abe","normalized_unit_value":"493/1000","observation_id":"954b3bb3d889d2d8fa0f84b9a3223eaf","value":0.493},{"configuration":"gemini-3.8-flash_high","model_id":"m-422e5ba87859dd55","normalized_unit_value":"697/1000","observation_id":"f06072893cbefd23dc63a86432677691","value":0.697},{"configuration":"gpt-6-astra_max","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"189/250","observation_id":"3557cefe053ef3a2ca1c95595065295f","value":0.756},{"configuration":"glm-5.2_max","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"171/500","observation_id":"bcb4fb304c488f86e521408952f81841","value":0.342},{"configuration":"kimi-k2.7-code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"73/200","observation_id":"29624d33497b02671ed8ec8a79e7f2e8","value":0.365},{"configuration":"grok-4.7_xhigh","model_id":"m-7187cf2b01d41935","normalized_unit_value":"14/25","observation_id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","value":0.56},{"configuration":"mistral-large-4-0_high","model_id":"m-72b806dc67894cf3","normalized_unit_value":"1/5","observation_id":"39e767a0ed9e71374bab635fd9e2ce92","value":0.2},{"configuration":"gemini-3.6-flash_high","model_id":"m-77b5dbc317075e9d","normalized_unit_value":"331/500","observation_id":"8f6725165e10f2ce17308791e15f9d1c","value":0.662},{"configuration":"gemini-4-argon_high","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"19/25","observation_id":"56d131d2a5cbdffe166bdda67e598654","value":0.76},{"configuration":"qwen3.8-max-0902_xhigh","model_id":"m-825009614c016c1d","normalized_unit_value":"473/1000","observation_id":"dd26bef806a0bba267bbd1b02770b43f","value":0.473},{"configuration":"gemini-3.5-flash_high","model_id":"m-8d99798a7e24690c","normalized_unit_value":"331/500","observation_id":"f90868b4246f3217c1fa61fa9b95b789","value":0.662},{"configuration":"claude-fable-5-1_max","model_id":"m-98ee7005cf45b532","normalized_unit_value":"177/250","observation_id":"add598a99258de2861924fc25c825623","value":0.708},{"configuration":"deepseek-v4.1-flash_max","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"429/1000","observation_id":"66b5a46f01ac8af0a7586871cb050ba1","value":0.429},{"configuration":"qwen3.8-max_xhigh","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"229/500","observation_id":"2e690ecc6ebeb390f087e351bd8c7803","value":0.458},{"configuration":"kimi-k3_max","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"253/500","observation_id":"63b449d94d8f27d0e405ea5ec8384866","value":0.506},{"configuration":"qwen3.7-max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"1115346038114343/2000000000000000","observation_id":"cb8036f627af27545731c0fdb0405c49","value":0.5576730190571715},{"configuration":"grok-4.5_high","model_id":"m-b0c41bb29e51489d","normalized_unit_value":"483/1000","observation_id":"5fb5dec681fa13f63b1b4d2e3c76490c","value":0.483},{"configuration":"Inkling_xhigh","model_id":"m-b439c804161fac26","normalized_unit_value":"403/1000","observation_id":"28411df2fb9e1660a30498ca8b0c9245","value":0.403},{"configuration":"glm-5.3_max","model_id":"m-c65e3e17492500e9","normalized_unit_value":"41/100","observation_id":"dd5c55d1f80c2ba62c676b7e3e150a64","value":0.41},{"configuration":"deepseek-v4.1-flash-1017_max","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"17/40","observation_id":"8e35d6850446c900d7c667bcebf2d5eb","value":0.425},{"configuration":"claude-opus-4-8_max","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"53/100","observation_id":"dbd9d582709dbd41201aa5f60842dfb0","value":0.53},{"configuration":"claude-opus-5-5_max","model_id":"m-cfe0652e9d685783","normalized_unit_value":"361/500","observation_id":"144b27b350bf50d4c19f78baa94c183a","value":0.722},{"configuration":"claude-sonnet-5-5_max","model_id":"m-d123610543e5f5ed","normalized_unit_value":"93/200","observation_id":"b5dfd8ffb656e8526bed64b86b168f2a","value":0.465},{"configuration":"Muse-Spark-Open-816B-A42B_xhigh","model_id":"m-db208dec31031420","normalized_unit_value":"283/500","observation_id":"2fc7380a8d1e0cbfdc6837142943acdd","value":0.566},{"configuration":"gpt-6-luna_max","model_id":"m-e5d076977e951607","normalized_unit_value":"207/500","observation_id":"d183987f58d811d89f8c42a38fe26df4","value":0.414},{"configuration":"deepseek-v4-pro-0813_max","model_id":"m-f936349cbdb1def9","normalized_unit_value":"5290581162324649/10000000000000000","observation_id":"665cc51403349d6e59eed00558808309","value":0.5290581162324649}],"protocol_id":"4a5d6edc10aebf40572422c5","unit":"比例"},"7ce5cd29eef0cb3a66875c099b7fba95051ef8fab4469549b5d7135a4ba1caf1":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"207/5","observation_id":"d28a3120853d4efe5840e7bb51950fc3","value":41.4},{"configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"513/10","observation_id":"b60faaa3ac13688ab0e4d3540dcbbf95","value":51.3},{"configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","model_id":"m-98ee7005cf45b532","normalized_unit_value":"157/5","observation_id":"92973184577a9ca58d3865a6bc131f18","value":31.4},{"configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","model_id":"m-cfe0652e9d685783","normalized_unit_value":"85/2","observation_id":"c4b221a2cb6a2006aac48962128f37ff","value":42.5},{"configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"539/10","observation_id":"3bdbfd4c6e52ba4e0916b5c06403b6c3","value":53.9}],"protocol_id":"f658b6663efcee3dc42343bd","unit":"%"},"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3":{"benchmark":"MMMU-Pro","benchmark_version":"Standard 10","members":[{"configuration":"GPT5.6-Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"83","observation_id":"723e50c3bda3d52daa8d895fc55553b6","value":83.0},{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"79","observation_id":"e6e7f5a00f3abe3c013f4beacb96537c","value":79.0},{"configuration":"Fable5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"406/5","observation_id":"a6ccc017ec5976d43545f1578c4ba3f8","value":81.2},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"823/10","observation_id":"d1f2e18c925df69cd229259da568a231","value":82.3},{"configuration":"Opus4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"378/5","observation_id":"d48870e5e71cd2cc3b689552b141ee4a","value":75.6}],"protocol_id":"bcb65796e6a9f26048c645d7","unit":"%"},"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"Claude Sonnet 5","model_id":"m-2cd37409a483435a","normalized_unit_value":"54","observation_id":"494d4b4ce09501b6feb0dbb80233b707","value":54.0},{"configuration":"GPT-5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"67","observation_id":"7ef9b74eae323671beacf43bf25de9bf","value":67.0},{"configuration":"Gemini 3.6 Flash","model_id":"m-77b5dbc317075e9d","normalized_unit_value":"49","observation_id":"2125ffdac9125e27a313037f91e5edb9","value":49.0},{"configuration":"Gemini 3.5 Flash","model_id":"m-8d99798a7e24690c","normalized_unit_value":"37","observation_id":"7b14fb9c23111ece0c4ecf6bf486054c","value":37.0},{"configuration":"Grok 4.5","model_id":"m-b0c41bb29e51489d","normalized_unit_value":"54","observation_id":"f4858ed8f9e084fde242d77a18dddeed","value":54.0}],"protocol_id":"7ed0e78158569ef8175dbb6a","unit":"%"},"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"473/5","observation_id":"06022591029d9808b7b2c118e22277aa","value":94.6},{"configuration":"Gemini 3.8 Flash","model_id":"m-422e5ba87859dd55","normalized_unit_value":"953/10","observation_id":"4523e0388c826e5440921d454fffec5e","value":95.3},{"configuration":"GPT‑6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"96","observation_id":"e40148b85734dc0f9f8ce08ccb902b8d","value":96.0},{"configuration":"Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"937/10","observation_id":"e80c9d5d961020a2775a86f5a8e8c7ff","value":93.7},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"923/10","observation_id":"710e7d489e0ed28e0e46334dd5f83f89","value":92.3},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"463/5","observation_id":"b6943c0f15a31bf97046c2e4af0b90f3","value":92.6},{"configuration":"Claude Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"937/10","observation_id":"7d3842228f84e6a5e57e7db96b8839e3","value":93.7},{"configuration":"Claude Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"92","observation_id":"29bbe2be4ccedeb019b3ccf6a3404eb9","value":92.0},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"929/10","observation_id":"6b87188fa1e95a23261b2a5e267b3413","value":92.9}],"protocol_id":"032545e2298934b8fc66f83f","unit":"%"},"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"MiniCPM-V 4.6 1.3B","model_id":"m-0cd6b9e2d04f2c30","normalized_unit_value":"61010101010101/200000000000000","observation_id":"c0db3282b4d173e68108771ffab82d19","value":0.305050505050505},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"558585858585859/1000000000000000","observation_id":"7e93a07b82dfd690e1c2b1309992332e","value":0.558585858585859},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"513131313131313/1000000000000000","observation_id":"249b24c9d6325f053552fb911f7f4a84","value":0.513131313131313},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"631313131313131/1000000000000000","observation_id":"80f85d325d6536dbd1603ecf4678090b","value":0.631313131313131},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"856565656565657/1000000000000000","observation_id":"7c25d00a19511ccc369639cf5605291e","value":0.856565656565657},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"86489898989899/125000000000000","observation_id":"c96afdd76392eb5d2c9b157b68b7de9e","value":0.691919191919192},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"756565656565657/1000000000000000","observation_id":"eebaf08b0e483d418dfe3796053cb2a9","value":0.756565656565657},{"configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"169090909090909/200000000000000","observation_id":"012394f027c8418fe233e5f26b558306","value":0.845454545454545},{"configuration":"JT-35B-Flash","model_id":"m-2d9c4a1754ad2072","normalized_unit_value":"829292929292929/1000000000000000","observation_id":"3c0c826ad87b9bae6195614ce7295dbf","value":0.829292929292929},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"178989898989899/200000000000000","observation_id":"70d4757947a89f4f143b905fc208d5b4","value":0.894949494949495},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"823232323232323/1000000000000000","observation_id":"4ac219e06b407a4da2953a84f21a360f","value":0.823232323232323},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"351010101010101/500000000000000","observation_id":"d95e6a6f385be987c8949ca6952efa66","value":0.702020202020202},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"952525252525253/1000000000000000","observation_id":"93a25d2afe298e1250966bfe70f4f70f","value":0.952525252525253},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"960606060606061/1000000000000000","observation_id":"c863d641570942861efdba9c19093efa","value":0.960606060606061},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"748484848484849/1000000000000000","observation_id":"f6a22fbde90853e88aebc327f5973f62","value":0.748484848484849},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"223989898989899/250000000000000","observation_id":"7516363fe155a8bbdfc789cee4300773","value":0.895959595959596},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"913131313131313/1000000000000000","observation_id":"185f07372e1def6a3525e6b1b6c635f5","value":0.913131313131313},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"38989898989899/50000000000000","observation_id":"7c995d56b5fa39eef927afa2cc879ab9","value":0.77979797979798},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"367171717171717/500000000000000","observation_id":"3e2102f0b8440ab308d1e2fedeb7c1ef","value":0.734343434343434},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"22885101010101/31250000000000","observation_id":"b5ca8a579b13fb2adc96c2c993971f7c","value":0.732323232323232},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"9/10","observation_id":"b8284820623630cb7a5150e367b7bed4","value":0.9},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"430808080808081/500000000000000","observation_id":"c2e927593d58bea9e466eb1677b5bc62","value":0.861616161616162},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"890909090909091/1000000000000000","observation_id":"8e38fd38f9044aff514003f73b248716","value":0.890909090909091},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"631313131313131/1000000000000000","observation_id":"63aced451500a12a72f3e068bcd6ff1f","value":0.631313131313131},{"configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","model_id":"m-7b2a8fe101678ebd","normalized_unit_value":"733333333333333/1000000000000000","observation_id":"bbee1bae6aa788cf47294ddc38f9761f","value":0.733333333333333},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"232070707070707/250000000000000","observation_id":"65e0dc8726d7a3e7afde224768d43963","value":0.928282828282828},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"760606060606061/1000000000000000","observation_id":"13075042d3067db5be44e3af38eac3de","value":0.760606060606061},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"292929292929293/1000000000000000","observation_id":"990a7e694606915f66b15d0a96c06abb","value":0.292929292929293},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"161111111111111/250000000000000","observation_id":"651c19da954966c3a71036c8d3db174d","value":0.644444444444444},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"937373737373737/1000000000000000","observation_id":"0539c349124105a728b2ffaeadc2dabd","value":0.937373737373737},{"configuration":"G9v3-3B","model_id":"m-9bafaddca6a68b50","normalized_unit_value":"219191919191919/500000000000000","observation_id":"d0469bd1e07d662d47e3eb56a0937741","value":0.438383838383838},{"configuration":"Motif 3","model_id":"m-9e1e0f605d47cf02","normalized_unit_value":"417171717171717/500000000000000","observation_id":"8763ad1982a06ca4a70e322ffe1ddbd4","value":0.834343434343434},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"743434343434343/1000000000000000","observation_id":"02699c8c436679b49b6319aa832d4d14","value":0.743434343434343},{"configuration":"Gemma 4 12B (Reasoning)","model_id":"m-a4be17f49ae915ce","normalized_unit_value":"188131313131313/250000000000000","observation_id":"6fa185e6f3a52003678d56e35ea14de0","value":0.752525252525252},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"411111111111111/500000000000000","observation_id":"da90bd5db2e44d989ddf563c8ec38c6b","value":0.822222222222222},{"configuration":"MiniCPM5-1B (Reasoning)","model_id":"m-ab31824c5d59e70f","normalized_unit_value":"138888888888889/500000000000000","observation_id":"9d5afdcfe1870948e303c2bcd8e408b7","value":0.277777777777778},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"866666666666667/1000000000000000","observation_id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","value":0.866666666666667},{"configuration":"G9v3-39A5B","model_id":"m-af0edba2f76c80d6","normalized_unit_value":"161010101010101/200000000000000","observation_id":"29813bdc52dc98722342b740c2f2d3f2","value":0.805050505050505},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"217929292929293/250000000000000","observation_id":"a95cdd62235bf41ff0b78b6a5a866d88","value":0.871717171717172},{"configuration":"Solar Open2 250B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"856565656565657/1000000000000000","observation_id":"f0716aa67008d67c9923512a4d8c5e11","value":0.856565656565657},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"167070707070707/200000000000000","observation_id":"ff2b5baedd439ff80001794125500940","value":0.835353535353535},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"111489898989899/125000000000000","observation_id":"004e3b6a13ca6828e88e6f4e76a97e76","value":0.891919191919192},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"829292929292929/1000000000000000","observation_id":"c948234960090e8c9ef801d4c2e468fc","value":0.829292929292929},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"188131313131313/250000000000000","observation_id":"885831acf33b9d6286fb559cc7b4ec0e","value":0.752525252525252},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"917171717171717/1000000000000000","observation_id":"85f2af62bad122fb48d79f2a5cd0ba76","value":0.917171717171717},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"181010101010101/200000000000000","observation_id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","value":0.905050505050505},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"228030303030303/250000000000000","observation_id":"3551c30b97ba4cb5f0fc94e087d94b59","value":0.912121212121212},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"215909090909091/250000000000000","observation_id":"90978c171b499bb2efc92c2da0d369b4","value":0.863636363636364},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"278787878787879/500000000000000","observation_id":"c0382c8dd0c506272b3fef2eaf1a7bcf","value":0.557575757575758},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"896969696969697/1000000000000000","observation_id":"9b6a2e4e945c90d8754fc862b2e6383b","value":0.896969696969697},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"872727272727273/1000000000000000","observation_id":"b588de5f182f53797aeaba57de63e608","value":0.872727272727273},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"856565656565657/1000000000000000","observation_id":"6d2464487e35f420dae74891c6e0dec0","value":0.856565656565657},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"929292929292929/1000000000000000","observation_id":"67157f1183905451cf4dded1391fab8e","value":0.929292929292929},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"88989898989899/100000000000000","observation_id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","value":0.88989898989899},{"configuration":"Agnes 2.5 Pro Beta","model_id":"m-e060cf24e4176269","normalized_unit_value":"181010101010101/200000000000000","observation_id":"2f5e3b6ebf2112f4f734c9c632e248bf","value":0.905050505050505},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"923232323232323/1000000000000000","observation_id":"c6f5e59b8c457f8fd73cf7fe23e60786","value":0.923232323232323},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"419191919191919/500000000000000","observation_id":"f63fe95853d236ed53d2b79d02b859cb","value":0.838383838383838},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"232070707070707/250000000000000","observation_id":"a4b58fb51111dbcc6b81f75ff3f7fc61","value":0.928282828282828},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"187070707070707/200000000000000","observation_id":"b2e2268e4647b383f382e5a7860dafa3","value":0.935353535353535},{"configuration":"DiffusionGemma 26B A4B","model_id":"m-fe0a26f770e1e011","normalized_unit_value":"668686868686869/1000000000000000","observation_id":"9804205b8a1d620728568c184381a38c","value":0.668686868686869}],"protocol_id":"ac37418879bd94f391def2ae","unit":"比例"},"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b":{"benchmark":"Chartography","benchmark_version":"current published task suite","members":[{"configuration":"Haiku 5.5","model_id":"m-016da3c7828e329e","normalized_unit_value":"232/5","observation_id":"888dd585273ea6255f0bebeb11182076","value":46.4},{"configuration":"Sonnet 5.5 For reference","model_id":"m-d123610543e5f5ed","normalized_unit_value":"308/5","observation_id":"34ec823cb4bb85ff7965822a48a3f4ce","value":61.6},{"configuration":"GPT-6 Luna","model_id":"m-e5d076977e951607","normalized_unit_value":"291/10","observation_id":"4a70f456ec596dc9670786d2fcc37f1a","value":29.1}],"protocol_id":"8ee4ee2a2791528dd623c2bb","unit":"%"},"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"GPT-6 Sol (max)","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"344/5","observation_id":"0c33bab4ff416863111d8ddbb4eb7f0b","value":68.8},{"configuration":"GPT-6 Luna (max)","model_id":"m-e5d076977e951607","normalized_unit_value":"333/5","observation_id":"12d12a117fed600435582eb0cdccab38","value":66.6}],"protocol_id":"2c719e8ae7701b7ffdc132b5","unit":"%"},"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1":{"benchmark":"HLE","benchmark_version":"Diamond","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"37/125","observation_id":"de9339207cccf727d5f73de793318c18","value":0.296},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"147/250","observation_id":"9165499121e5e198448d4536365d9c16","value":0.588},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"29/1000","observation_id":"20cab07f9ffc43544259f41d417e323c","value":0.029},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"1/25","observation_id":"1c035cc3ccd2789e7031fcc995e521f7","value":0.04},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"473/1000","observation_id":"e85b4ef3cccfd0ec6841edff67807576","value":0.473},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"97/200","observation_id":"4e26762c5adc39cb8336491e6db427a8","value":0.485},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"21/500","observation_id":"19f9062134531b120e5a3beb1b827183","value":0.042},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"121/1000","observation_id":"3c34699bd1ded8eb4f916044a6a266ef","value":0.121},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"11/125","observation_id":"30beb0844cc09771243cbe9dfff243cd","value":0.088},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"73/500","observation_id":"5706c727b7c4831ffe377f7ebcc17796","value":0.146},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"19/500","observation_id":"75182b4b31b9383c53f3f18d0e09c65d","value":0.038},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"331/1000","observation_id":"0ec8a0a14f14573543e2e6dd2102670c","value":0.331},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"319/500","observation_id":"e829d764b6c12d2fb9cdd8b27d5bf1fe","value":0.638},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"47/1000","observation_id":"5c0267c76de42807699083ce1e0581c4","value":0.047},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"37/250","observation_id":"7c461dbccbd2115e299061bd3cfb0f10","value":0.148},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"137/1000","observation_id":"433cbb1c2fcf649c36b004a451515cf8","value":0.137},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"211/1000","observation_id":"c42c9d1f70b755ec31ff5dcd6c984f87","value":0.211},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"373/1000","observation_id":"8cb18c97abf2e38058caf77fed29461c","value":0.373},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"39/100","observation_id":"c7cb147aa61e00d80a403c5843ea98f0","value":0.39},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"23/200","observation_id":"12e6c7bd6a0e1d4d1047be66dc2b360a","value":0.115},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"131/1000","observation_id":"92b0eaaad67a910391a173c1c1816701","value":0.131},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"29/125","observation_id":"353c470ec7dd55d7fbbc616285180c31","value":0.232},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"139/1000","observation_id":"00361f2a88ddce87fbff29f4d32f2681","value":0.139},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"19/250","observation_id":"4fb23f5b07fa697e50fa64d17b095add","value":0.076},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"173/1000","observation_id":"3c62c893b3f2b3d94df10089539fb20a","value":0.173},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"47/1000","observation_id":"66e35510159ab65c6e04c83322148b65","value":0.047},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"57/1000","observation_id":"04be87f5b8e53109685297a0e0936bc5","value":0.057},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"567/1000","observation_id":"b185a4fb3502b3f94e9e93582e0d93a3","value":0.567},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"6/25","observation_id":"2ab50d9a5a2d86ccb853771d429875d3","value":0.24},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"69/1000","observation_id":"5c428eec7c6c7eddbfd00d536e4c24e0","value":0.069},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"51/1000","observation_id":"48d08d1a4f13b1f74f0a43f0e70c78cb","value":0.051},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"281/1000","observation_id":"a1e07609863e9a13d3798af4a816cb48","value":0.281},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"43/1000","observation_id":"db2ce17572462c1eb76d15ea72c97e2b","value":0.043},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"27/500","observation_id":"11ffdf294b4e40190b29f471a44eb232","value":0.054},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"601/1000","observation_id":"6cd57d6545a81a30a933d08c69e8d902","value":0.601},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"9/200","observation_id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","value":0.045},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"43/250","observation_id":"39803a6fda08687f289513743c809091","value":0.172},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"17/200","observation_id":"85bd1df1c636ca5159d1bed6e1a0f499","value":0.085},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"53/500","observation_id":"2d291e3811085b2648de857339509203","value":0.106},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"16/125","observation_id":"8319710b72494f93151d3337f0dbcdfd","value":0.128},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"327/1000","observation_id":"5198f9fefccca8b5961dab3ccd43cfb2","value":0.327},{"configuration":"Mercury 2.5","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"51/1000","observation_id":"c6b63609fd0ed9a9d1cc3300620f0885","value":0.051},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"91/1000","observation_id":"67c349c9897445ce219d7f8acc0e4dbe","value":0.091},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"57/1000","observation_id":"1e3489d548cb804c3cfe4ff52df120e8","value":0.057},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"179/1000","observation_id":"20084307dcc6cac0bc9c76758b5e2139","value":0.179},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"153/1000","observation_id":"ee3f7405021072ff698c63b6e1fa510f","value":0.153},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"253/1000","observation_id":"2bfa42798a174eb5a88b71f065cd4aef","value":0.253},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"91/500","observation_id":"a57cb2704c93293d339e91362f40ae46","value":0.182},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"679/1000","observation_id":"3522855b1814574fd5b3733b32b77769","value":0.679},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"29/250","observation_id":"7efb7f319954edefb37b62ce3d3ff826","value":0.116},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"533/1000","observation_id":"c57acd914b7734768f36308b43cd96b9","value":0.533},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"29/250","observation_id":"a0be97a286389603a66382ac4d794e81","value":0.116},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"51/500","observation_id":"03d73659cb7b3abbbd4921c8546a8418","value":0.102},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"4/25","observation_id":"4912807a823c5e919553ca48ade2b67a","value":0.16},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"19/125","observation_id":"2e68084b0515fdc6a19fe2b7a59814d2","value":0.152},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"263/1000","observation_id":"440f5ca7aab0d03094cae5ee8558ca3d","value":0.263},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"691/1000","observation_id":"2541b171f3a9b70175c5eff4ce5a2a47","value":0.691},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"163/1000","observation_id":"1e856649fb09c4da785f8c4ea8ad05e0","value":0.163},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"111/500","observation_id":"ca6b622f3e2324e23758731c3eedb79c","value":0.222},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"99/1000","observation_id":"cda0d514d2e746d7e415df4602a30e61","value":0.099},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"287/1000","observation_id":"89d939473fcae282b052740fc3fcfd7a","value":0.287},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"243/1000","observation_id":"ea28428a53ddb22639d7f5b7e50547a2","value":0.243},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"39/200","observation_id":"d543f8764a78abfabdc05d4f312b8b2f","value":0.195},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"201/1000","observation_id":"b6295a8854af11f8a1045539e4932330","value":0.201}],"protocol_id":"09ebba97991aa218642551d3","unit":"比例"},"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e":{"benchmark":"APEX-Agents","benchmark_version":"current published task suite","members":[{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"28023598820059/125000000000000","observation_id":"274b775cdefb96d48e1e4a5a4e4f52de","value":0.224188790560472},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"424041297935103/1000000000000000","observation_id":"ae1b88ab2d0880e90cfd75002e9f4d77","value":0.424041297935103},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"274336283185841/1000000000000000","observation_id":"ca1ee026c3875e1f8a6f67a549e74fe8","value":0.274336283185841},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"311946902654867/1000000000000000","observation_id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","value":0.311946902654867},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"94948377581121/250000000000000","observation_id":"519c33c19b165c6197391af117e522e3","value":0.379793510324484}],"protocol_id":"c455abb49f64fc5efd099eb8","unit":"比例"},"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"腾讯官方比较表 · GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"399/10","observation_id":"fadaedd3ee1f675326bbb9970170cf61","value":39.9},{"configuration":"腾讯官方比较表 · Hy4 preview","model_id":"m-383f40728e3e5259","normalized_unit_value":"321/10","observation_id":"f68fab6bf158dd2600b8426c509f20a3","value":32.1},{"configuration":"腾讯官方比较表 · Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"487/10","observation_id":"773afe0f7c5b5cfa8f123e1e37bca353","value":48.7},{"configuration":"腾讯官方比较表 · Qwen3.8 Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"411/10","observation_id":"37b17217723693963901c38560492d58","value":41.1},{"configuration":"腾讯官方比较表 · Kimi K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"91/2","observation_id":"9d73635ebbe34cac1967dc999b07ad7a","value":45.5},{"configuration":"腾讯官方比较表 · GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"247/5","observation_id":"083544e7fa726861ec84d82ed3ea9450","value":49.4},{"configuration":"腾讯官方比较表 · Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"161/10","observation_id":"ea9839b4879e46422a95dd0f7e029b6a","value":16.1},{"configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","model_id":"m-f936349cbdb1def9","normalized_unit_value":"152/5","observation_id":"7aef04ad579d81beb4f5c8ec2832a80c","value":30.4}],"protocol_id":"0c0199cb037df16b91698054","unit":"%"},"8d9a0e59927d81a1315e8322e2266a6c683a91285cd4c7fbd03e2785b92e1a9b":{"benchmark":"Terminal-Bench","benchmark_version":"4.0","members":[{"configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"591/10","observation_id":"109d26a2b8ea6ca9e4a3d9999b7c5ce6","value":59.1},{"configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"571/10","observation_id":"39d29b69788dd7332642aeadf008dc34","value":57.1},{"configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","model_id":"m-98ee7005cf45b532","normalized_unit_value":"52","observation_id":"fc1ff1f2174461295bf4f514896624b4","value":52.0},{"configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","model_id":"m-cfe0652e9d685783","normalized_unit_value":"298/5","observation_id":"bbe3b0185e63421d7e8bf29fa803e68e","value":59.6},{"configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"326/5","observation_id":"01c0338617b2ef52cd797f5b878bbe71","value":65.2}],"protocol_id":"6dfc6fa538421f2fee6d35c4","unit":"%"},"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844":{"benchmark":"IFBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT 5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"727/10","observation_id":"e818a4b1e85e949ea72540313ee99c3c","value":72.7},{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"73","observation_id":"a7d850e0b3d6527713784858f859aa3f","value":73.0},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"51","observation_id":"9a125a53ca950d873e4577684bf386d1","value":51.0},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"663/10","observation_id":"30e5129b448d30519f98acdcb4661feb","value":66.3},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"127/2","observation_id":"93d9797cec6ccd384ba684d52f6088bc","value":63.5},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"414/5","observation_id":"71646c5fb8bcced91b2a83b600993b46","value":82.8},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"791/10","observation_id":"68966a1d85f137636ad7fa5c27b6e2d8","value":79.1},{"configuration":"Solar Open 2 250B-A15B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"80","observation_id":"dafb6731eb5f8725711540078af7013b","value":80.0},{"configuration":"Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"311/5","observation_id":"0a5ea9e2a9d5f26a6a034bb3853216a5","value":62.2},{"configuration":"LFM2.5-2.6B (2.6B)","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"5917/100","observation_id":"49b7d39b3c4a3eaca06d3883d0335282","value":59.17}],"protocol_id":"81d4b10b8d99c4374542c3ff","unit":"%"},"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"229/5","observation_id":"b1f2d23ce08087c377528f7267508376","value":45.8},{"configuration":"DS-V4-Flash","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"377/10","observation_id":"25c531e52c90edae0ad06f8953351733","value":37.7},{"configuration":"Opus-5.0","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"503/10","observation_id":"6fa1a31f17a7792b499d20a7e15bd3ed","value":50.3},{"configuration":"DS-V4.1-Flash","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"274/5","observation_id":"e28fafc89bbfaddac23f5b76027e9f84","value":54.8},{"configuration":"K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"467/10","observation_id":"ab9a351b7071ab7f587fbaf6f7ef4c51","value":46.7},{"configuration":"GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"244/5","observation_id":"28ec00bcc7a9b0d7e2be777a2deb9b0e","value":48.8},{"configuration":"DS-V4-Pro","model_id":"m-f936349cbdb1def9","normalized_unit_value":"216/5","observation_id":"b7e9e3548b08b017e44f0d19beea862d","value":43.2}],"protocol_id":"9312268968d4ddf4fd457042","unit":"%"},"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093":{"benchmark":"OSWorld","benchmark_version":"2.1","members":[{"configuration":"Haiku 5.5","model_id":"m-016da3c7828e329e","normalized_unit_value":"362/5","observation_id":"5b785287811a33de601a7060b80bab3e","value":72.4},{"configuration":"Sonnet 5","model_id":"m-2cd37409a483435a","normalized_unit_value":"57","observation_id":"cdb864bc0c75bd1fb1f8c43c173d544c","value":57.0},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"74","observation_id":"5f71dd3fff3f2fe7b85716b5a982127d","value":74.0},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"807/10","observation_id":"520b3d1c3a49b9ec75dd20571c69ae96","value":80.7},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"409/5","observation_id":"cb71933d15a49daa0088c6b64854af7f","value":81.8},{"configuration":"Sonnet 5.5 For reference","model_id":"m-d123610543e5f5ed","normalized_unit_value":"839/10","observation_id":"e0c59ec207d44193213f25c93cf18f5b","value":83.9},{"configuration":"GPT-6 Luna","model_id":"m-e5d076977e951607","normalized_unit_value":"489/10","observation_id":"b4bc853a9a856396a7313b82125218ee","value":48.9}],"protocol_id":"e6a07dd8a64cf51e658fe61b","unit":"%"},"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7":{"benchmark":"FrontierMath","benchmark_version":"2","members":[{"configuration":"claude-haiku-5-5_max","model_id":"m-016da3c7828e329e","normalized_unit_value":"926829268292683/2000000000000000","observation_id":"21e96013106dd82d62904afe0c4f768b","value":0.4634146341463415},{"configuration":"gpt-6.1-sol_max","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"1","observation_id":"cacd9412ff6ba035ea9e826693b09343","value":1.0},{"configuration":"gpt-6.1-terra_max","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"926829268292683/1000000000000000","observation_id":"428bcb6f2312f54ea131660ec5510f0c","value":0.926829268292683},{"configuration":"muse-wonder-1.0_max","model_id":"m-68480fc0423c5034","normalized_unit_value":"7073170731707317/10000000000000000","observation_id":"a21e66eb62f2bfca72e7c625965b008a","value":0.7073170731707317},{"configuration":"mistral-large-4-0_high","model_id":"m-72b806dc67894cf3","normalized_unit_value":"2682926829268293/10000000000000000","observation_id":"8462c2a29fbbd735e478ce804db0a935","value":0.2682926829268293},{"configuration":"gemini-4-argon_high","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"975609756097561/1250000000000000","observation_id":"1004f40613fd7110bf20d6c47c5dcedb","value":0.7804878048780488},{"configuration":"deepseek-v4.1-flash_max","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"2682926829268293/10000000000000000","observation_id":"3121b8aa930fa12d352b2910814de607","value":0.2682926829268293},{"configuration":"deepseek-v4.1-flash-1017_max","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"926829268292683/2000000000000000","observation_id":"f2e8485b9221333706b0e9e14cd8c4e7","value":0.4634146341463415},{"configuration":"claude-sonnet-5-5_max","model_id":"m-d123610543e5f5ed","normalized_unit_value":"8048780487804879/10000000000000000","observation_id":"a0152968ddf4e79954f370988a3aa877","value":0.8048780487804879},{"configuration":"Muse-Spark-Open-816B-A42B_xhigh","model_id":"m-db208dec31031420","normalized_unit_value":"3170731707317073/10000000000000000","observation_id":"c2edb36eb77a36c2b7e3b5065aafdd4c","value":0.3170731707317073}],"protocol_id":"df0b5ec86237f8f4a709ac95","unit":"比例"},"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"腾讯官方比较表 · GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"689/10","observation_id":"b7c2bea8fb6f0a31dd7c2532db1fe168","value":68.9},{"configuration":"腾讯官方比较表 · Hy4 preview","model_id":"m-383f40728e3e5259","normalized_unit_value":"643/10","observation_id":"49c0ea386ce0f4eee60fb157edec7d0d","value":64.3},{"configuration":"腾讯官方比较表 · Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"747/10","observation_id":"2cc0820ff84711ce0c2bfadbfbb013aa","value":74.7},{"configuration":"腾讯官方比较表 · Qwen3.8 Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"278/5","observation_id":"6ff1592097b1d5537d18b6c552613651","value":55.6},{"configuration":"腾讯官方比较表 · Kimi K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"74","observation_id":"9a67d273dfa56310e03c5e50dc85cc6e","value":74.0},{"configuration":"腾讯官方比较表 · GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"681/10","observation_id":"a1f523e170da53cbcf5ad3cbcb817e6a","value":68.1},{"configuration":"腾讯官方比较表 · Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"28","observation_id":"6f7ac5ece00998e78042608639ac6e80","value":28.0},{"configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","model_id":"m-f936349cbdb1def9","normalized_unit_value":"294/5","observation_id":"981ac64200af40a5afa217a11e6600b2","value":58.8}],"protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","unit":"%"},"9bfe8c7aec98429e18deb5dbb07961818a8e0a1acdf8c4d30bab0b9e84f1353d":{"benchmark":"Terminal-Bench-Science","benchmark_version":"0.1.0","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"112/5","observation_id":"b48576f95bc7cd0aae9b966cadaca0c7","value":22.4},{"configuration":"GPT‑6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"323/5","observation_id":"c754fd67237feab08d16e123eec898ee","value":64.6},{"configuration":"Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"30","observation_id":"d8372588b76cfe9aa6be7425ce13cfb0","value":30.0},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"107/5","observation_id":"fe5b288d983212a01455a7bed99bcbe3","value":21.4},{"configuration":"Claude Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"263/5","observation_id":"e9ea1a0c9a1ad3ceb806b63de3f0f7a8","value":52.6}],"protocol_id":"6e315c8a4731951529a013b5","unit":"%"},"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b":{"benchmark":"EnterpriseOps-Gym","benchmark_version":"current published task suite","members":[{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"1067591763652641/2500000000000000","observation_id":"fe3cd8a1335112d2ed6891e37698119e","value":0.4270367054610564},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"3372127723067741/10000000000000000","observation_id":"da0d77fdb30d1e419d687691f9d0af7b","value":0.3372127723067741},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"10056699492688749/25000000000000000","observation_id":"1d959564327d6d5496bf9174d774a4c8","value":0.40226797970754996},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"4709042076991943/10000000000000000","observation_id":"54f26e677fb3a678985d92f416cf28ad","value":0.4709042076991943},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"40555058191584603/100000000000000000","observation_id":"e323a2908ddfad0f233fccef67ea0b29","value":0.40555058191584603},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"4762757385854969/10000000000000000","observation_id":"aaa4039880a2e3c56a0fee65a0cf496e","value":0.4762757385854969},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"18352730528200537/100000000000000000","observation_id":"80ae66f2b7393fe1f51bab7846dd0a10","value":0.18352730528200537},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"1155475977320203/4000000000000000","observation_id":"4c1c120149f7031df34c6f3ac8ff9629","value":0.28886899433005075},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"37958818263205013/100000000000000000","observation_id":"40747d0e01fbdeb26f3ad425860f13bb","value":0.37958818263205013},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"34735899731423453/100000000000000000","observation_id":"6751eebce76aacc484338eb2b0237376","value":0.34735899731423453},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"36436884512085943/100000000000000000","observation_id":"4a39f12346421fa0779ea6e76813ff8f","value":0.36436884512085943},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"4422560429722471/10000000000000000","observation_id":"cea12af74b5609df87befba4292e8ab2","value":0.4422560429722471},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"3321396598030439/10000000000000000","observation_id":"b8b704019523524af0ce859e89927fa0","value":0.3321396598030439},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"3210981796478663/10000000000000000","observation_id":"df3f56d8282f0ce2c7d297df530d227a","value":0.3210981796478663},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"105863921217547/250000000000000","observation_id":"5ffca0a973b63389e8c0ae797941349a","value":0.423455684870188},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"468218442256043/1000000000000000","observation_id":"ccd3047c248ced9697a972d3c394d08d","value":0.468218442256043},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"2358997314234557/5000000000000000","observation_id":"6be4e7acd5ffa2c31291fe47e51a2a60","value":0.4717994628469114},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"4959713518352731/10000000000000000","observation_id":"6c40f85770ab1fe03b45de2938f913b3","value":0.4959713518352731},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"147996866606983/312500000000000","observation_id":"32758e4ec324e0ec55fbc6f952cf8937","value":0.4735899731423456}],"protocol_id":"b19a5badb616725bc5909b39","unit":"比例"},"a33a54887ef6afdde922d0df84c343493c2a6ff51326312daeb375ead7374d3a":{"benchmark":"Terminal-Bench-Science","benchmark_version":"0.1.0","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"112/5","observation_id":"fad33698f014ebb1ac938c69d7c0798b","value":22.4},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"29","observation_id":"74d5382d9b4d1ccd5a786120f5a550f0","value":29.0},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"247/10","observation_id":"34ae78b3ccf7f2237d26759361270f8d","value":24.7},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"263/5","observation_id":"fb4474bb9b423ef25754804c5a39561f","value":52.6}],"protocol_id":"88023c6718be09c275d82cfc","unit":"%"},"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2":{"benchmark":"AIME","benchmark_version":"2026","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"167/2","observation_id":"9bcefe5681f1de0d56a93cb5ca7a6411","value":83.5},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"567/10","observation_id":"072b96c67ae19a20ee48a971117995c0","value":56.7},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"173/2","observation_id":"0f70d96aa863b344234e9f6fb914e9f2","value":86.5},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"226/5","observation_id":"1aa7620d5715919ff1d95bb835bf5bb1","value":45.2}],"protocol_id":"97974a29c366bdc315df37df","unit":"%"},"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"GPT-5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"73","observation_id":"6817c08660542a8ba2cec888e8e593bf","value":73.0},{"configuration":"GLM-5.2 (max)","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"231/5","observation_id":"546ef51508bcfe588737e01c240c5015","value":46.2},{"configuration":"Claude Fable 5 (max, w/ fallback)","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"70","observation_id":"f8d14a3afa2d20c0373188e9c415dc6c","value":70.0},{"configuration":"Kimi K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"135/2","observation_id":"c87699e8afdc2f67ab86923665dd9ed5","value":67.5},{"configuration":"Claude Opus 4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"59","observation_id":"04a252c77bb5bb362fedd335d947b332","value":59.0}],"protocol_id":"3b7e071f01bc65c78bc74f4a","unit":"%"},"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d":{"benchmark":"Terminal-Bench","benchmark_version":"4.0","members":[{"configuration":"Claude Haiku 5.5 / Mini-SWE-agent","model_id":"m-016da3c7828e329e","normalized_unit_value":"909/25","observation_id":"e2ba5f9af59c0998db11f28273d765a7","value":36.36},{"configuration":"GPT-5.6 Sol / Mini-SWE-agent","model_id":"m-0510680735d3c79a","normalized_unit_value":"947/25","observation_id":"340dbce590b6aaaa8e1dc907ce4c05ef","value":37.88},{"configuration":"GPT-6.1 Sol / Mini-SWE-agent","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"1101/20","observation_id":"d8d7274643c06b1a4ca112516b5792d9","value":55.05},{"configuration":"Gemini 3.7 Flash / Mini-SWE-agent","model_id":"m-16a8dce2a6e7167a","normalized_unit_value":"303/25","observation_id":"d7c99e26bef72e8691defabf85dff136","value":12.12},{"configuration":"GPT-6.1 Terra / Mini-SWE-agent","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"4949/100","observation_id":"6b7dbc78828ec370d1445c253fb081f8","value":49.49},{"configuration":"Muse Wonder Full / Mini-SWE-agent","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"1452/25","observation_id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","value":58.08},{"configuration":"Claude Sonnet 5 / Mini-SWE-agent","model_id":"m-2cd37409a483435a","normalized_unit_value":"1061/100","observation_id":"bbf181bad55ef3d8ecad294914ed6b67","value":10.61},{"configuration":"Muse Spark 1.2 / Mini-SWE-agent","model_id":"m-2f22b07da3af863c","normalized_unit_value":"303/50","observation_id":"bd6d00a622ee8666fe845f21e2f0650f","value":6.06},{"configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"1869/100","observation_id":"6bb0c331de392cf1dc35a1de5de30317","value":18.69},{"configuration":"Inkling Small / Mini-SWE-agent","model_id":"m-35c4f287acc76c95","normalized_unit_value":"0","observation_id":"2f300304d9d00cf74194cf028a1e2a05","value":0.0},{"configuration":"Hy4 preview / Mini-SWE-agent","model_id":"m-383f40728e3e5259","normalized_unit_value":"202/25","observation_id":"54a707812fad1f7aeac89c3fb45c1b29","value":8.08},{"configuration":"GPT-6 Sol / Mini-SWE-agent","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"1111/25","observation_id":"a142db33093dc7d6f056b8d8509f429e","value":44.44},{"configuration":"Grok 4.6 / Mini-SWE-agent","model_id":"m-3c17c0d3ca7f1abe","normalized_unit_value":"1717/100","observation_id":"620e1ee85f651d96e2f6e21d676ecc23","value":17.17},{"configuration":"Gemini 3.8 Flash / Mini-SWE-agent","model_id":"m-422e5ba87859dd55","normalized_unit_value":"197/10","observation_id":"ed92f5e3f147fe079ea8a940470282f6","value":19.7},{"configuration":"GPT-6 Astra / Mini-SWE-agent","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"298/5","observation_id":"28b064c611914ab9c6c8b9c66db69c4f","value":59.6},{"configuration":"Claude Opus 5 / Mini-SWE-agent","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"5353/100","observation_id":"7df1cc76a5ce426695809152ab7c2b62","value":53.53},{"configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","model_id":"m-60cd64beda366768","normalized_unit_value":"606/25","observation_id":"3fbf327bd9cbc7d6e3195a9d909b9231","value":24.24},{"configuration":"GPT-5.6 Luna / Mini-SWE-agent","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"581/50","observation_id":"5bec3b19db8e7a3362bcd865766b2cc8","value":11.62},{"configuration":"Kimi K3.1 Pro / Mini-SWE-agent","model_id":"m-66c1aae520b7679a","normalized_unit_value":"1591/25","observation_id":"952e1d48294b90de64278faf65544562","value":63.64},{"configuration":"Muse Wonder / Mini-SWE-agent","model_id":"m-68480fc0423c5034","normalized_unit_value":"4949/100","observation_id":"cbbd47b78fc1abe8ff45d7f2b54da87d","value":49.49},{"configuration":"Grok 4.7 / Mini-SWE-agent","model_id":"m-7187cf2b01d41935","normalized_unit_value":"2879/100","observation_id":"bcf918003f716a22bcd694aa5c29a3d6","value":28.79},{"configuration":"Gemini 4 Argon / Mini-SWE-agent","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"2879/50","observation_id":"e3cc310052e96951792e7addae519033","value":57.58},{"configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","model_id":"m-825009614c016c1d","normalized_unit_value":"3889/100","observation_id":"0d2e487f06bf23cbe7ee7f90f6003d16","value":38.89},{"configuration":"Claude Fable 5 / Mini-SWE-agent","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"4141/100","observation_id":"5ecd3796946e727694a868a88027022a","value":41.41},{"configuration":"Gemini 3.5 Flash / Mini-SWE-agent","model_id":"m-8d99798a7e24690c","normalized_unit_value":"303/50","observation_id":"349a8da908aa998fbc2c641cf58cfc1b","value":6.06},{"configuration":"MiniMax-M3.1-Flash / Mini-SWE-agent","model_id":"m-8e76831a5814eec9","normalized_unit_value":"101/4","observation_id":"3aab4755f21529c39094e459048b22d6","value":25.25},{"configuration":"Claude Fable 5.1 / Mini-SWE-agent","model_id":"m-98ee7005cf45b532","normalized_unit_value":"1452/25","observation_id":"cd6feaf81208453ed0051fee8e3b0d5a","value":58.08},{"configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"197/10","observation_id":"4ce54d74c2045140751a930d5f81e63e","value":19.7},{"configuration":"Qwen3.8 Max / Mini-SWE-agent","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"1717/50","observation_id":"e085a37b9093e9f4b5222ae848bc472f","value":34.34},{"configuration":"Kimi K3 / Mini-SWE-agent","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"1717/100","observation_id":"fff5c31ca1adab72967ff7c824d4653e","value":17.17},{"configuration":"Grok 4.5 / Mini-SWE-agent","model_id":"m-b0c41bb29e51489d","normalized_unit_value":"859/100","observation_id":"a49b610cdf02973ee5519a3d58738257","value":8.59},{"configuration":"Inkling / Mini-SWE-agent","model_id":"m-b439c804161fac26","normalized_unit_value":"51/100","observation_id":"ea8212294d42fc5eee3233e729cf83c0","value":0.51},{"configuration":"Beam 1.0 / Mini-SWE-agent","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"101/25","observation_id":"82df891de38afd61fe2c16983a26faa8","value":4.04},{"configuration":"Mercury 2.5 / Mini-SWE-agent","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"0","observation_id":"3f9c90b35f6b5c160dd7d713ceef5f66","value":0.0},{"configuration":"Muse Spark 1.3 / Mini-SWE-agent","model_id":"m-c0580b01221ed260","normalized_unit_value":"1061/100","observation_id":"73167c475c92b2f6b68a011a68a996ab","value":10.61},{"configuration":"GLM-5.3 / Mini-SWE-agent","model_id":"m-c65e3e17492500e9","normalized_unit_value":"3889/100","observation_id":"379465a1dce0780ec05eea7fd360c800","value":38.89},{"configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"1591/50","observation_id":"6d06cabd5082651a9ea6ff6f7709c136","value":31.82},{"configuration":"Claude Opus 4.8 / Mini-SWE-agent","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"2323/100","observation_id":"2539da8c839981606943a4efe846b325","value":23.23},{"configuration":"GLM-5.3-Flash / Mini-SWE-agent","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"3737/100","observation_id":"28abab0f2a19e622cf15ffcdc9be3dfd","value":37.37},{"configuration":"Claude Opus 5.5 / Mini-SWE-agent","model_id":"m-cfe0652e9d685783","normalized_unit_value":"1303/20","observation_id":"bab68b4a8cd2763b38ac011956432e06","value":65.15},{"configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","model_id":"m-d123610543e5f5ed","normalized_unit_value":"3207/50","observation_id":"886c25f56626744a389c5cae640c1a98","value":64.14},{"configuration":"MiniMax-M3 / Mini-SWE-agent","model_id":"m-da421acad97d607b","normalized_unit_value":"101/100","observation_id":"8f361f3f1d062dfe3e449864fa84aec9","value":1.01},{"configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"3131/100","observation_id":"65da79b527ae8b650ca14c9d0ac7489d","value":31.31},{"configuration":"Muse Spark Open / Mini-SWE-agent","model_id":"m-db208dec31031420","normalized_unit_value":"657/100","observation_id":"8362f29198914b33fdad64ba36fb745e","value":6.57},{"configuration":"Gemini 4 Carbon / Mini-SWE-agent","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"6313/100","observation_id":"2db4ed4fb97d84958b935e83df5db43d","value":63.13},{"configuration":"GPT-6 Luna / Mini-SWE-agent","model_id":"m-e5d076977e951607","normalized_unit_value":"341/25","observation_id":"a8bde16affe89631785f57ec46935865","value":13.64},{"configuration":"Step 5 Preview / Mini-SWE-agent","model_id":"m-f892623396612d9b","normalized_unit_value":"1591/50","observation_id":"d5a9a67230f3e72676fe2415a004dd47","value":31.82},{"configuration":"GPT-5.6 Terra / Mini-SWE-agent","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"2273/100","observation_id":"7fbb0050443eb76fe54b1727ad2dc1fe","value":22.73},{"configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","model_id":"m-f936349cbdb1def9","normalized_unit_value":"707/50","observation_id":"35687651705594abdbf62890a0bd5a87","value":14.14}],"protocol_id":"dbbbcaf695179b55af0fc57c","unit":"%"},"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e":{"benchmark":"FrontierMath","benchmark_version":"2","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"89","observation_id":"5b5d809739c95968ceacfb8058855082","value":89.0},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"393/5","observation_id":"0c9a231c5ad369432b2d4e6968b6bc7c","value":78.6},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"87","observation_id":"663672da9a7ca9303639232126d54075","value":87.0},{"configuration":"Claude Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"80","observation_id":"8bd910e44b3e0be43d3313d301ed4e94","value":80.0},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"849/10","observation_id":"34d11ef45eddcdc63e92f205eb9eaa3c","value":84.9}],"protocol_id":"3842bb51683d99f4ac9c82c3","unit":"%"},"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad":{"benchmark":"DeepSWE","benchmark_version":"1.1","members":[{"configuration":"GPT5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"73","observation_id":"fa113390898a4649fa21eed1a4997857","value":73.0},{"configuration":"Fable5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"70","observation_id":"99378a4787b7b2fe45ecccbd3c678fc6","value":70.0},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"283/5","observation_id":"9619717553c9af1d9fdaa3ac01eac67f","value":56.6},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"108/5","observation_id":"abbe21aa88bf4a572f15ca2233e791d6","value":21.6},{"configuration":"Opus4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"59","observation_id":"5b6ba99bf6e1deff4ca257f908136e39","value":59.0}],"protocol_id":"eb23265b5e5c19deb4dff9c1","unit":"%"},"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9":{"benchmark":"Multi-IF","benchmark_version":"current maintainer dataset","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"759/10","observation_id":"cf2e1f5dc39c9244f47a835c7f33569d","value":75.9},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"357/5","observation_id":"5b52e92a8499178ddfa84c801132f2f4","value":71.4},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"359/5","observation_id":"ca424296b6739f797da253012aa25cc9","value":71.8},{"configuration":"LFM2.5-2.6B (2.6B)","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"8007/100","observation_id":"ddcaace3bb0b4a1452f5dea314dd058c","value":80.07}],"protocol_id":"9400ddb5b35b4ffd1dad74b3","unit":"%"},"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"181/10","observation_id":"29e83d6eda99c0b01ee6e4a2561a6131","value":18.1},{"configuration":"GPT‑6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"207/5","observation_id":"583c9d1e2da082747c42402376e2e5f7","value":41.4},{"configuration":"Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"269/10","observation_id":"e54e313a4cf3bafac7d6e3d2385d1025","value":26.9},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"149/10","observation_id":"51dfce6bb3b25136b0ea3545031f03b8","value":14.9},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"87/5","observation_id":"c935b5fe18707c971521047b772f70d1","value":17.4},{"configuration":"Gemini 3.5 Flash","model_id":"m-8d99798a7e24690c","normalized_unit_value":"29/2","observation_id":"d209f552e4c2026bb776833fb9d113ca","value":14.5},{"configuration":"Claude Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"157/5","observation_id":"2372d74bdc501d7556cf0aa74944c192","value":31.4},{"configuration":"Claude Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"31/2","observation_id":"6bd66300b839f4ff34d5fd8249350449","value":15.5},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"76/5","observation_id":"e4ce479712f2cad43715215ba9603696","value":15.2}],"protocol_id":"a93de83dd8cfc372c687b91d","unit":"%"},"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82":{"benchmark":"MMMU-Pro","benchmark_version":"Standard 10","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"423/5","observation_id":"267266c10eb3ebd28421d19f9b4fb8da","value":84.6},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"159/2","observation_id":"bc79c1e1910e14e0ff6ab222415c13af","value":79.5},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"82","observation_id":"079f06724159c79071056bf02b730573","value":82.0}],"protocol_id":"f5c96106a212a39d97b2e63c","unit":"%"},"ba65a0607782d37e37ba8c548a0ffc6603c65b39eb5bd71828b7dc8322062c5e":{"benchmark":"Terminal-Bench","benchmark_version":"4.0","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"399/10","observation_id":"6ee2b87b3178b7a9ed8fa714c84647bf","value":39.9},{"configuration":"DS-V4-Flash","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"7","observation_id":"cb8519de408cafe4f74e189f2d0ff6f2","value":7.0},{"configuration":"Opus-5.0","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"259/5","observation_id":"ac8f6dd5c815b356a6b57350a5fb1c7e","value":51.8},{"configuration":"DS-V4.1-Flash","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"156/5","observation_id":"9641c48c69c00cab23b6294d207c8f62","value":31.2},{"configuration":"K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"63/5","observation_id":"f3a815210c20a49420a2227fcd8d8f19","value":12.6},{"configuration":"GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"379/10","observation_id":"9a912208deecc3da383c31d8417739d7","value":37.9},{"configuration":"DS-V4-Pro","model_id":"m-f936349cbdb1def9","normalized_unit_value":"62/5","observation_id":"8c629283bcf934e5a97ae61bcef9fec0","value":12.4}],"protocol_id":"d3512b6f8cf5dcb54718109d","unit":"%"},"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223":{"benchmark":"BFCL","benchmark_version":"4","members":[{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"729/10","observation_id":"355d178390cc7668a17845fc6539c179","value":72.9},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"75","observation_id":"0886e98ce0984c34f09c65e2184f8692","value":75.0}],"protocol_id":"a643d03be3b50ea260b8cf17","unit":"%"},"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b":{"benchmark":"Chartography","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"799/10","observation_id":"c10b5e0559e0dd2ee1b376a9b0908819","value":79.9},{"configuration":"Opus-5.0","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"84","observation_id":"c1b1f70a9c86ed29fb6aed62d8d0a04b","value":84.0},{"configuration":"DS-V4.1-Flash","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"789/10","observation_id":"e5a64d88a17b8e69890219a69e61c957","value":78.9},{"configuration":"K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"681/10","observation_id":"24019ec9f1eb6cf5c33ae9562d1f1bac","value":68.1}],"protocol_id":"5aad08fc7a196bcc6b2b847f","unit":"%"},"c1099eb5410d7c4dca7f212922ce69d84c268751da0920c2e04b6542c51bea89":{"benchmark":"Terminal-Bench","benchmark_version":"4.0","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"373/10","observation_id":"3a2ab6999fa584f73c8d3dab12d76e7c","value":37.3},{"configuration":"Gemini 3.8 Flash","model_id":"m-422e5ba87859dd55","normalized_unit_value":"191/10","observation_id":"60b632021bc45635455840f54b94c4f4","value":19.1},{"configuration":"GPT‑6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"579/10","observation_id":"3817ce25a9a6b0599640f57a9cca3571","value":57.9},{"configuration":"Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"263/5","observation_id":"7db5563eacfc0ff2de1106fa741f24cd","value":52.6},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"89/2","observation_id":"be49aacf10e50a43b90ee123c1c07cdd","value":44.5},{"configuration":"Claude Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"279/5","observation_id":"ceb85c397387a4847b40dd8b3d454d6e","value":55.8}],"protocol_id":"ac0f80ad7501f371bea4ff3e","unit":"%"},"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3":{"benchmark":"AA-Briefcase","benchmark_version":"1.1","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"39353/25","observation_id":"6fc39f46271fccbf5e9ec9022c726f0f","value":1574.12},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"155509/100","observation_id":"c7f025daa6b0e96ab52dc7df7e3bb96a","value":1555.09},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"9883/100","observation_id":"caf840132854029b414465a52c19889d","value":98.83},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"31041/100","observation_id":"2e097f3022508687e5c06477d01a5253","value":310.41},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"35034/25","observation_id":"07c0ef81ba075e768f25e1177f4ce283","value":1401.36},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"43233/25","observation_id":"4f4f6c3a22e75b41412d7850bc4b9cf5","value":1729.32},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"13159/20","observation_id":"d8f9d30bc51b5ac382358dbda4a533e5","value":657.95},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"79297/100","observation_id":"ed41f889fc4e7786b137bd6b76a3da51","value":792.97},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"4249/20","observation_id":"f3f72292d6ff516b21ceaec7da17a866","value":212.45},{"configuration":"JT-4.1 Flash 236B A21B (Reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"27311/25","observation_id":"6f61748806659fcf7e221477d42b6563","value":1092.44},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"91263/100","observation_id":"b87f8c47a158b30537b60e5453c5b4d9","value":912.63},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"24969/25","observation_id":"b0a10dcacf57ed9fc64b8506d23f5ae1","value":998.76},{"configuration":"Nemotron 3.5 Bolt 74B-A9B","model_id":"m-3ad11f2b6253b364","normalized_unit_value":"45597/50","observation_id":"e9c29563b5cc25446296aa38a0ecd4ae","value":911.94},{"configuration":"GPT-5.5 Instant (June 2026)","model_id":"m-3fe24456aa661a33","normalized_unit_value":"54333/50","observation_id":"110052f3248350cfda8c6580116d3dd9","value":1086.66},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"42329/100","observation_id":"8a20412103b559527bf3c1a6c26511e5","value":423.29},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"120299/100","observation_id":"c6c37db45460e80aac980c7255d8ffed","value":1202.99},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"156859/100","observation_id":"e3f8cbf64c84e82bb4a86c035247a89a","value":1568.59},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"26163/50","observation_id":"671c80eaf4d486bab2d265a1c67045fd","value":523.26},{"configuration":"Kimi K2.7 Code","model_id":"m-5ea285ac7903cbc7","normalized_unit_value":"21451/25","observation_id":"8c6b16e6477c30b02c0cf413d3d98aff","value":858.04},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"148683/100","observation_id":"e5699451024628001ea794ad1191793b","value":1486.83},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"7092/5","observation_id":"94b4404155c32f58241674a265734f95","value":1418.4},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"167417/100","observation_id":"2b512abdb78eff5062c8cbfe38f0a197","value":1674.17},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"33173/20","observation_id":"849c220b415c509e507a4751daf48b0a","value":1658.65},{"configuration":"LongCat 2.0","model_id":"m-6ac02cf910375655","normalized_unit_value":"73973/100","observation_id":"1e5075fa69ac15167616fbc24b2c2ddf","value":739.73},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"47591/100","observation_id":"662e05367e55e07efb079daf8cc94493","value":475.91},{"configuration":"Quasar 438B (Max, Based on GLM-5.2)","model_id":"m-71070d6294f8dc57","normalized_unit_value":"22471/20","observation_id":"f0265d9e7d722340c4dcae3d534155b9","value":1123.55},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"8187/5","observation_id":"7ce9f4dd8dbbce093592c7a8603e6868","value":1637.4},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"18307/20","observation_id":"397f726491d251a65df1f043233be5de","value":915.35},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"19699/20","observation_id":"3a94cd016eb68403ef51f517cb4b5722","value":984.95},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"134831/100","observation_id":"bc12c3c00877323848c26ea730d935cc","value":1348.31},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"27333/25","observation_id":"1caa339c8531983b8c2298b6a2b6d27c","value":1093.32},{"configuration":"Celeris-1","model_id":"m-7990b74b056a0b9a","normalized_unit_value":"7597/50","observation_id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","value":151.94},{"configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","model_id":"m-7b2a8fe101678ebd","normalized_unit_value":"23241/100","observation_id":"606c6cc32d7931469129472a4d6fffcd","value":232.41},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"148503/100","observation_id":"782453050df6541e218363a18f9bb044","value":1485.03},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"80307/50","observation_id":"ecc6d7d48519d5a999fc5f71c76c0b6d","value":1606.14},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"29046/25","observation_id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","value":1161.84},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"1781/5","observation_id":"d623fd47b28047c5307a67919c5c6fae","value":356.2},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"73799/50","observation_id":"da58b11d089faf8c31c386386035abd7","value":1475.98},{"configuration":"K2 Horizon 0.9B","model_id":"m-928c410ba1082113","normalized_unit_value":"0","observation_id":"3f053e6933f1e5ef77b0bf87fdfb5476","value":0},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"1579/4","observation_id":"afdcb1efb0aca22bea9c16a3d016167d","value":394.75},{"configuration":"Ling-3.0-flash-Fin","model_id":"m-95ce1c534519e66a","normalized_unit_value":"95863/100","observation_id":"11eea946a9e871701b9adaec2c3d14b2","value":958.63},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"166949/100","observation_id":"cab25ff85e887187f7f900705800483c","value":1669.49},{"configuration":"Motif 3","model_id":"m-9e1e0f605d47cf02","normalized_unit_value":"3967/4","observation_id":"56f3576f6b7e7bada3d4ea81a3e974d2","value":991.75},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"10137/20","observation_id":"47de14e8ef5807902d31f4e62879bab3","value":506.85},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"1522","observation_id":"2d7e1c988db8f9f1d1a2378f0202de51","value":1522},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"116479/100","observation_id":"54e0dbe28b524361d2c39e892adaa185","value":1164.79},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"87709/100","observation_id":"e12061e69225a05e579dff8f6ab0f0de","value":877.09},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"16661/20","observation_id":"7d5fcb038f359de5556a57151b51cffc","value":833.05},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"4881/5","observation_id":"fb238aac3df6d482092b3981302c93d1","value":976.2},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"144669/100","observation_id":"732f112b308243fb337093d80efff0c3","value":1446.69},{"configuration":"Mercury 2.5","model_id":"m-b9d9f5989f59e750","normalized_unit_value":"37849/100","observation_id":"c4bef304832e9ad8f150cf2ae6ad2c77","value":378.49},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"47783/100","observation_id":"6b81d5a0603a102ca44c8276f19acede","value":477.83},{"configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","model_id":"m-c0755395b2da9721","normalized_unit_value":"47013/50","observation_id":"82a245d0d5111b9661f660b2a94b3db6","value":940.26},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"65109/100","observation_id":"42d65fa891e8c3a4f35b4ea031798999","value":651.09},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"101351/100","observation_id":"6d3a1ea10b80de3baf1922e2ef5fc22e","value":1013.51},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"37571/25","observation_id":"98953845b6332ba7948a61f814a750d3","value":1502.84},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"34784/25","observation_id":"bee797f9fafa487f710b7abf6c3fddcf","value":1391.36},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"78521/50","observation_id":"e9014188fe35aeb589dfba302eb5ab58","value":1570.42},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"72547/50","observation_id":"3015c7ad3cd2d55a621f41dcebd98f3d","value":1450.94},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"179989/100","observation_id":"5c16c97705a8f70907dc1ba381949448","value":1799.89},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"25304/25","observation_id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","value":1012.16},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"181333/100","observation_id":"609c8649b3a04b44b0d6807fdfe15427","value":1813.33},{"configuration":"Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"23681/25","observation_id":"a684e0a09f0e331ba14d29191ea44f68","value":947.24},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"32312/25","observation_id":"4c2ed1cac020a530bd05f7d661a8c39c","value":1292.48},{"configuration":"A.X-K2","model_id":"m-d48cde87e1f9c28d","normalized_unit_value":"73917/100","observation_id":"628eb864f68ddd810e173dc996c88c01","value":739.17},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"108909/100","observation_id":"703518382f29bb0a80ddba827b81e22d","value":1089.09},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"37797/25","observation_id":"c6fb2c23c770d5e7bda4e380c2517fe7","value":1511.88},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"127869/100","observation_id":"00f62480bdc3ab7dea00bbda96ff9331","value":1278.69},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"161063/100","observation_id":"fe408037e14234a002d5a2d7a3b186f0","value":1610.63},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"157351/100","observation_id":"62ba5fa448ec9e7825ee5093f17e487d","value":1573.51},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"66853/50","observation_id":"27c5dd58443826f50a40adf5e5fbdf05","value":1337.06},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"65117/100","observation_id":"0ec16dc40db0414f2891438008ea0a1a","value":651.17},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"76343/50","observation_id":"ae0bf141fcba5d508b0da173dcf01538","value":1526.86},{"configuration":"Ling 3.1 Flash","model_id":"m-f74476b1532b3a4f","normalized_unit_value":"69847/50","observation_id":"d6fbe929d324cc028c3c012f923590ba","value":1396.94},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"71259/50","observation_id":"5342274720726a814f5f5d822f07b510","value":1425.18},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"31329/25","observation_id":"40fe84ac7f2d6389219a5c06f59da083","value":1253.16},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"43387/50","observation_id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","value":867.74},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"35759/25","observation_id":"be21483d79b65c8a6ad01a6eee1e0103","value":1430.36}],"protocol_id":"e0d5f43b6db1c39db024366a","unit":"分"},"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68":{"benchmark":"FrontierMath","benchmark_version":"2","members":[{"configuration":"GPT‑5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"83","observation_id":"2e3935359cea7e94134e77b2b6c16a83","value":83.0},{"configuration":"GPT‑6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"488/5","observation_id":"0f175b01de79f7689d7f750dac02d89b","value":97.6},{"configuration":"Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"366/5","observation_id":"88ff5cf09a8ea0336d12c5e7aac222d3","value":73.2},{"configuration":"GPT‑5.6 Luna","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"117/2","observation_id":"ee6e51ce0e1c80e8c9e718c1bdb9fdb9","value":58.5},{"configuration":"Claude Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"451/5","observation_id":"51cb1dea03ece9a899f843aedba7597e","value":90.2},{"configuration":"Claude Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"439/5","observation_id":"a450c91f279ff648cb2b208a74552ba0","value":87.8},{"configuration":"Claude Opus 4.8","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"561/10","observation_id":"040f48951e0ebb9ff6bfbde83c816f48","value":56.1},{"configuration":"GPT‑5.6 Terra","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"683/10","observation_id":"dde4f14f270dc23fa94482cb2b80d4e8","value":68.3}],"protocol_id":"8d02db73f46a4d9d6e92f159","unit":"%"},"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb":{"benchmark":"AutomationBench","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"297/10","observation_id":"35d3b6c21baca0074691fdf7baa39656","value":29.7},{"configuration":"GLM-5.2 (max)","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"129/10","observation_id":"96fb24279652563e7d2bc82b71d0fc52","value":12.9},{"configuration":"Claude Fable 5 (max, w/ fallback)","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"291/10","observation_id":"127d7af68dabb544e18741601786e437","value":29.1},{"configuration":"Kimi K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"154/5","observation_id":"cf571d5463813d8daec30e8a8f31740d","value":30.8},{"configuration":"Claude Opus 4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"136/5","observation_id":"682b2a95186a2a4bc27f60c8b3158f30","value":27.2}],"protocol_id":"afcaaccdf9ce24d5aa463683","unit":"%"},"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1":{"benchmark":"BFCL","benchmark_version":"4","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"261/5","observation_id":"c9019a0cb657ee535803619a061368cf","value":52.2},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"246/5","observation_id":"61ff3d8dc14ad46a6ba4cce95c802615","value":49.2},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"333/5","observation_id":"9fed535a16c1cf8e07c033c97e3a6783","value":66.6},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"611/10","observation_id":"79e538fa5b2bc837bbbc3882037e1347","value":61.1}],"protocol_id":"add40720ac2f4db6b3eb0fb3","unit":"%"},"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"GPT 5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"941/10","observation_id":"593853e47efc92fb5fb50665ad18805a","value":94.1},{"configuration":"GLM-5.2 (max)","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"456/5","observation_id":"55c9779a030f53f08065b75a89097776","value":91.2},{"configuration":"Fable 5","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"463/5","observation_id":"6fcca244b7e23fe3074831e7ca3cf1be","value":92.6},{"configuration":"Qwen3.8-Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"463/5","observation_id":"7ff900665f3a2edd6bd47eeb54dcc81e","value":92.6},{"configuration":"Kimi K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"187/2","observation_id":"dd3025c1130213f154e327d038efe0ed","value":93.5},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"462/5","observation_id":"a6934af3407d37537eb6eb1bc8d0e969","value":92.4},{"configuration":"Claude Opus 4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"91","observation_id":"b730b9240ece9fc9ae2c909c6d9b5b27","value":91.0}],"protocol_id":"920ae7339db73f9e73534a08","unit":"%"},"d92832fa55e82132b49a2c173a59ef8eb7a5a65696389a70f135715ba60ca73b":{"benchmark":"Terminal-Bench-Science","benchmark_version":"0.1.0","members":[{"configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"633/10","observation_id":"11d354c28bc62a914ec4cd9b63089852","value":63.3},{"configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"529/10","observation_id":"58e1da3fabc235dff575e99c81331c7e","value":52.9},{"configuration":"Claude Fable 5.1 / official Gemini 4 Carbon comparison","model_id":"m-98ee7005cf45b532","normalized_unit_value":"433/10","observation_id":"11ef98afa95070c53ee2bc064c7c7936","value":43.3},{"configuration":"Claude Opus 5.5 / official Gemini 4 Carbon comparison","model_id":"m-cfe0652e9d685783","normalized_unit_value":"59","observation_id":"47678bec4f89a1e5999bbadc4fe96b61","value":59.0},{"configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"324/5","observation_id":"77b00ca561419071138d999c141114f3","value":64.8}],"protocol_id":"4f191cd5eb9caf57e48bc0e7","unit":"%"},"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c":{"benchmark":"APEX-Agents","benchmark_version":"current published task suite","members":[{"configuration":"GPT-5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"399/10","observation_id":"62c8c73bb0229cb8c08fb44bb8f568ca","value":39.9},{"configuration":"GLM-5.2 (max)","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"178/5","observation_id":"6f3782a0a7f13ec12e5eae491c79c282","value":35.6},{"configuration":"Claude Fable 5 (max, w/ fallback)","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"433/10","observation_id":"b82b11faf1fed2434465112eb548316c","value":43.3},{"configuration":"Kimi K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"41","observation_id":"e4b1483bcb889a5ef567c4c02fc571dd","value":41.0},{"configuration":"Solar Open 2 250B-A15B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"83/5","observation_id":"01aeec9c08a057635f9b72825df4f3fb","value":16.6},{"configuration":"Claude Opus 4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"197/5","observation_id":"9584a53c71e055a6a275ffd8975b0b25","value":39.4}],"protocol_id":"427a1a2a441195d7f27614d9","unit":"%"},"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2":{"benchmark":"AA-Omniscience","benchmark_version":"1.1","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"3466666666666667/10000000000000000","observation_id":"30bc019f7dc3e7314870de193c23616d","value":0.3466666666666667},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"6183333333333333/10000000000000000","observation_id":"ad20786a83f6b0b25b2f134cb2f6c095","value":0.6183333333333333},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"23/250","observation_id":"432ed3bb705952b8c4dcd7c697936b37","value":0.092},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"11766666666666667/100000000000000000","observation_id":"4a7b9f5169709891b5917ee76e2a8e1a","value":0.11766666666666667},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"47633333333333333/100000000000000000","observation_id":"43cafc16779c4f9a017fe8ae9ccdcb81","value":0.47633333333333333},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"4096666666666667/10000000000000000","observation_id":"a92d6d21339d4ee97a6fa90820847da1","value":0.4096666666666667},{"configuration":"Ring-2.6-1T","model_id":"m-25ccaf97cf82a5e0","normalized_unit_value":"263/1000","observation_id":"f5ca62e67f09a75d5d460c81d0851fa4","value":0.263},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"18366666666666667/100000000000000000","observation_id":"2d5bfdd87d275a4c6177037eb5a3ff85","value":0.18366666666666667},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"159/500","observation_id":"c31ed6f5932783f10dbc8e7e910353cc","value":0.318},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"21166666666666667/100000000000000000","observation_id":"41481c594a0d6acdcc3cd6076cd49938","value":0.21166666666666667},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"5053333333333333/10000000000000000","observation_id":"1cb384dc9fec91c5db3c1841d018dfc3","value":0.5053333333333333},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"6156666666666667/10000000000000000","observation_id":"eec9e765997481a4684eb7ff4710087d","value":0.6156666666666667},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"23033333333333333/100000000000000000","observation_id":"2b5b8d191f3e5e3bfa651cee2bbbd819","value":0.23033333333333333},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"34/125","observation_id":"084eb35df8b2a6c0ecbd814a3709a769","value":0.272},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"24466666666666667/50000000000000000","observation_id":"e4c0067c2ce56fc8c2df02c769685194","value":0.48933333333333334},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"4754166666666667/12500000000000000","observation_id":"3f07a1a0073c74e060ddc96a3f9e8bcb","value":0.38033333333333336},{"configuration":"Ling 3.0 Tiny","model_id":"m-6b5176ed51518e35","normalized_unit_value":"4741666666666667/12500000000000000","observation_id":"97a4da3275fb19d0e062184a2de4a918","value":0.37933333333333336},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"427/1000","observation_id":"86bdedd227d21984b4d30c6ee2bdc8b4","value":0.427},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"13/50","observation_id":"6656d06e85a16beb97db67e4960f17c9","value":0.26},{"configuration":"Solar Pro 4","model_id":"m-754df8b5922bf5da","normalized_unit_value":"4608333333333333/25000000000000000","observation_id":"fcae50a41df6fb08e7a7f301689104ab","value":0.18433333333333332},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"483/1000","observation_id":"0588c4d7908c2d87deb09a5a03cbe1c3","value":0.483},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"14766666666666667/50000000000000000","observation_id":"c42d799db980160b63c3bab31d80fa02","value":0.29533333333333334},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"11666666666666667/100000000000000000","observation_id":"464c122b5f2d778a04114dc266e1a6b8","value":0.11666666666666667},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"9833333333333333/100000000000000000","observation_id":"0ca7e9728f633cc599fad0ed12a6097b","value":0.09833333333333333},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"22533333333333333/50000000000000000","observation_id":"2ab90fc472453676cb1037409906c906","value":0.45066666666666666},{"configuration":"Granite 4.2 30B","model_id":"m-9453d55ff407aaee","normalized_unit_value":"11233333333333333/100000000000000000","observation_id":"d19df36b3cd3a0e840c1256cd1c744fe","value":0.11233333333333333},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"171/250","observation_id":"b829940da89808936844666d294a78aa","value":0.684},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"7366666666666667/50000000000000000","observation_id":"8698ac189bdac0fe4a6655a8ed1ade73","value":0.14733333333333334},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"2653333333333333/10000000000000000","observation_id":"568a15734298d304532a5f2a4188d632","value":0.2653333333333333},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"4246666666666667/20000000000000000","observation_id":"7a931d4238635511e69e1feb35bfd055","value":0.21233333333333335},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"393/1000","observation_id":"95037c24ac83fd0eadfbf39605ac569f","value":0.393},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"6146666666666667/20000000000000000","observation_id":"11dc653c29e2dbe67eaa7a540042fef3","value":0.30733333333333335},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"9993333333333333/20000000000000000","observation_id":"795265b8e61a4f96a23b56d372ad2050","value":0.49966666666666665},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"2796666666666667/10000000000000000","observation_id":"ce91d9d68242cca613b5b932f9a344d4","value":0.2796666666666667},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"29/200","observation_id":"6b5784505206d57d0bba2ceb5b8de557","value":0.145},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"3233333333333333/10000000000000000","observation_id":"be51199fdc7d1731045d427201852660","value":0.3233333333333333},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"173/1000","observation_id":"5ce6a4a59546a3ac7bd2230a498f7f48","value":0.173},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"99/250","observation_id":"28843797b0f403bc00b3b129a0250f86","value":0.396},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"141/500","observation_id":"a76a9fe9ce50e17717765fcaebd6fbf8","value":0.282},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"131/200","observation_id":"3247fbdae3f488e2395593b52951cc4b","value":0.655},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"5146666666666667/10000000000000000","observation_id":"911853403e469f557e0c742924a5bacb","value":0.5146666666666667},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"4658333333333333/25000000000000000","observation_id":"d414f5fe5dd8b0e7a11df833233b1edf","value":0.18633333333333332},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"22433333333333333/50000000000000000","observation_id":"253af28f0e697d86e6b8e62f78eae062","value":0.44866666666666666},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"87/250","observation_id":"d99f9179103c8156a9cedc7180b05c42","value":0.348},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"21016666666666667/50000000000000000","observation_id":"dbcce4096cefcfd79cea60ad99bf1516","value":0.42033333333333334},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"5826666666666667/10000000000000000","observation_id":"00d59d147ff0b7cd3c12bdcaa9a4b516","value":0.5826666666666667},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"11766666666666667/50000000000000000","observation_id":"0351ad1258b35881ec0b312bb72c8d29","value":0.23533333333333334},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"3963333333333333/10000000000000000","observation_id":"efcb1058be3817d672ab8961f07089a0","value":0.3963333333333333},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"53/200","observation_id":"5ac3b36a409b29ee98389b739f2f25b7","value":0.265},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"33766666666666667/100000000000000000","observation_id":"28a0e72d64e34dbdaef952ff1ff713b4","value":0.33766666666666667},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"3636666666666667/10000000000000000","observation_id":"ec9021441721b216b9626f97b96062f1","value":0.3636666666666667},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"113/250","observation_id":"3f8c786b8fc874f5d0f0ae1174917198","value":0.452},{"configuration":"Solar Mini 4","model_id":"m-f99cd5ceed3a5b41","normalized_unit_value":"4558333333333333/25000000000000000","observation_id":"8e92630a337eaf823218ceb658e6daf8","value":0.18233333333333332},{"configuration":"Qwen3.8 2.4T A95B","model_id":"m-fb920161af1862b7","normalized_unit_value":"157/500","observation_id":"872cbb8006efe075a4c9b62b4f118011","value":0.314}],"protocol_id":"57ce78dfdf41b4d0acd6e37d","unit":"比例"},"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b":{"benchmark":"MMMU-Pro","benchmark_version":"Standard 10","members":[{"configuration":"MiniCPM-V 4.6 1.3B","model_id":"m-0cd6b9e2d04f2c30","normalized_unit_value":"379190751445087/1000000000000000","observation_id":"7a567c9ea42d069b6cee3f34d322f5a9","value":0.379190751445087},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"171907514450867/200000000000000","observation_id":"9a17a04ad8c7b9671f59a85153438a32","value":0.859537572254335},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"169364161849711/200000000000000","observation_id":"cd4afdb9196fc47218c2c1085364d08a","value":0.846820809248555},{"configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","model_id":"m-2d46f5cc91cc119d","normalized_unit_value":"320520231213873/500000000000000","observation_id":"90c2787f5b2a0845d431bc6af09c8c25","value":0.641040462427746},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"148092485549133/200000000000000","observation_id":"93ce1046100f7e346ed0a342303a4604","value":0.740462427745665},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"17121387283237/20000000000000","observation_id":"8bcd59ede7a9782bc9b8b3c169c19887","value":0.85606936416185},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"86878612716763/100000000000000","observation_id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","value":0.86878612716763},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"324277456647399/500000000000000","observation_id":"ce0faae345d00b20396f1024d947c26c","value":0.648554913294798},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"730635838150289/1000000000000000","observation_id":"f2f6f962e776bf31934ea15f004e8009","value":0.730635838150289},{"configuration":"DeepSeek V4 Flash Vision (Max)","model_id":"m-65378fc44ea2de26","normalized_unit_value":"747976878612717/1000000000000000","observation_id":"dbca8fa745f042a216711b700c5e06ca","value":0.747976878612717},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"419364161849711/500000000000000","observation_id":"75435cf97a5f0d98b110b2544004171f","value":0.838728323699422},{"configuration":"Muse Wonder (High)","model_id":"m-68480fc0423c5034","normalized_unit_value":"838150289017341/1000000000000000","observation_id":"bc79e5409c84a79a9897609e9d7ac22e","value":0.838150289017341},{"configuration":"Grok 4.7 (Low)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"78150289017341/100000000000000","observation_id":"6a15ad1c8de4bbfb578c4f789e53c9b8","value":0.78150289017341},{"configuration":"Qwen3.7 Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"804624277456647/1000000000000000","observation_id":"b6422caa44d095e13cde40b5c1a62bc3","value":0.804624277456647},{"configuration":"Ling-3.0-flash-VL","model_id":"m-724c2a784624afa2","normalized_unit_value":"789595375722543/1000000000000000","observation_id":"487c57fe9e6c668745daa6efed20c475","value":0.789595375722543},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"206936416184971/250000000000000","observation_id":"502735168b948696e5ae04c6357e58b6","value":0.827745664739884},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"158092485549133/250000000000000","observation_id":"b79c49748f47d291a424306d40dc6469","value":0.632369942196532},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"395664739884393/500000000000000","observation_id":"fc2b137ca09cf3cd0123e67faaeeed34","value":0.791329479768786},{"configuration":"Gemma 4 12B (Reasoning)","model_id":"m-a4be17f49ae915ce","normalized_unit_value":"348265895953757/500000000000000","observation_id":"ca0bc4ca214d2607b60ef7f92f73e72e","value":0.696531791907514},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"91835260115607/125000000000000","observation_id":"874b32395a8c924c1d3f0faf3d2e2b2e","value":0.734682080924856},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"743352601156069/1000000000000000","observation_id":"681af2b5c57d4ca46641c38951c3d479","value":0.743352601156069},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"763005780346821/1000000000000000","observation_id":"fa359632aae944ad7f900cc61aa55351","value":0.763005780346821},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"80635838150289/100000000000000","observation_id":"33026de8036ee6188f530a53e9aaa0aa","value":0.80635838150289},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"876878612716763/1000000000000000","observation_id":"3dabe34a660554a2d248f440d9ae144c","value":0.876878612716763},{"configuration":"Apodex 1.1","model_id":"m-d031501e2d0117d8","normalized_unit_value":"791907514450867/1000000000000000","observation_id":"d16dcb5ce825969190ea4f39de9f84b5","value":0.791907514450867},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"785549132947977/1000000000000000","observation_id":"9161d7aea787f48463e1dc1ca6952542","value":0.785549132947977},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"161849710982659/200000000000000","observation_id":"71c04ca8ce2b90585afee7d563834b34","value":0.809248554913295},{"configuration":"Gemini 4 Carbon (Medium)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"223121387283237/250000000000000","observation_id":"34112c44d835a7ef36f154c3c42aa2b1","value":0.892485549132948},{"configuration":"Agnes 2.5 Pro Beta","model_id":"m-e060cf24e4176269","normalized_unit_value":"754335260115607/1000000000000000","observation_id":"d7ea5990ff16dd400cf1a637a1392f4c","value":0.754335260115607},{"configuration":"Qwen3.8-Flash-Next","model_id":"m-e1cba225d1430824","normalized_unit_value":"199421965317919/250000000000000","observation_id":"42c55648590ad4a69cdc2a21de6faac7","value":0.797687861271676},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"398265895953757/500000000000000","observation_id":"7443158d8b6f105fcb7949451afd84ee","value":0.796531791907514},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"49385838150289/62500000000000","observation_id":"ef305cf290c211e130604e205ee97f46","value":0.790173410404624},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"163121387283237/200000000000000","observation_id":"e681be1aa8a9a82be96c894925b92cdd","value":0.815606936416185},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"381791907514451/500000000000000","observation_id":"e0d9cab1ce675bebaf2c47a0db9ddea9","value":0.763583815028902},{"configuration":"DiffusionGemma 26B A4B","model_id":"m-fe0a26f770e1e011","normalized_unit_value":"83164739884393/125000000000000","observation_id":"d47a048ab5b9e3796a7da71c7471a272","value":0.665317919075144}],"protocol_id":"9cc385f4bedc10cb3e53b89f","unit":"比例"},"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db":{"benchmark":"IFEval","benchmark_version":"current maintainer dataset","members":[{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"473/5","observation_id":"5ec142fbe451bd07f915ecd120f2faec","value":94.6},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"943/10","observation_id":"98b4a1e95c1679056cdd34a216b7399d","value":94.3}],"protocol_id":"25e5cdaaeea049303cac90a0","unit":"%"},"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde":{"benchmark":"AA-Briefcase","benchmark_version":"1.1","members":[{"configuration":"Haiku 5.5","model_id":"m-016da3c7828e329e","normalized_unit_value":"1578","observation_id":"f01303851f0ac5f8f962f37a067f5763","value":1578.0},{"configuration":"Sonnet 5","model_id":"m-2cd37409a483435a","normalized_unit_value":"1359","observation_id":"1454e68858e26e7f25431416e9f35b82","value":1359.0},{"configuration":"GPT-6 Sol","model_id":"m-3b5ea6fdfa94bc08","normalized_unit_value":"1483","observation_id":"ad451031d1d4a9ddfab5c379bfc5d852","value":1483.0},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"1822","observation_id":"56ec8a4eb3d885260337d38fdb4d7987","value":1822.0},{"configuration":"Sonnet 5.5 For reference","model_id":"m-d123610543e5f5ed","normalized_unit_value":"1824","observation_id":"be4f3c7647faa1a11f91719e70b95e94","value":1824.0},{"configuration":"GPT-6 Luna","model_id":"m-e5d076977e951607","normalized_unit_value":"1336","observation_id":"be25bcbd47f8306b2da2234ede387aa9","value":1336.0}],"protocol_id":"74db141f961ec29fa7ab8d6a","unit":"%"},"e7bc4f643ae0ec51ce57253962978c1d267f518628caeebcc42f9422b2fd9e21":{"benchmark":"Terminal-Bench-Science","benchmark_version":"0.1.0","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"112/5","observation_id":"cc26eea60ec9309fc3beabf4435c173d","value":22.4},{"configuration":"GPT-6 Astra","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"323/5","observation_id":"bf3d7c1e3fc2beed65427c4ff2fa2858","value":64.6},{"configuration":"Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"29","observation_id":"1f1de51f1003f9f9b7cc66b798316c8d","value":29.0},{"configuration":"Fable 5.1","model_id":"m-98ee7005cf45b532","normalized_unit_value":"263/5","observation_id":"0d7edb34eeb6882f692808525c3c2a73","value":52.6},{"configuration":"Opus 5.5","model_id":"m-cfe0652e9d685783","normalized_unit_value":"587/10","observation_id":"35aa2ee19acc44897fc2654a03403f3a","value":58.7}],"protocol_id":"1975ca8c09f87d3fc7597c12","unit":"%"},"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3":{"benchmark":"GPQA Diamond","benchmark_version":"Diamond","members":[{"configuration":"GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"941/10","observation_id":"a730c185c3223a52793bf1eef3941811","value":94.1},{"configuration":"DS-V4-Flash","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"899/10","observation_id":"021a82f4f7be85e4b693852ce483ef03","value":89.9},{"configuration":"Opus-5.0","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"467/5","observation_id":"d4f91d80df262b71392602c5e046be89","value":93.4},{"configuration":"DS-V4.1-Flash","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"909/10","observation_id":"6c2c7071889205aa8d51f6bb7f06fb84","value":90.9},{"configuration":"K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"929/10","observation_id":"233e224a230e5689366760183b9c92e8","value":92.9},{"configuration":"GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"881/10","observation_id":"4f28c1e35ee6e421b89e2a85fbc8ba80","value":88.1},{"configuration":"DS-V4-Pro","model_id":"m-f936349cbdb1def9","normalized_unit_value":"462/5","observation_id":"dc36fb77534eeb86099f253b7683ccaf","value":92.4}],"protocol_id":"f442e5a9ad3210e681dbc34e","unit":"%"},"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d":{"benchmark":"APEX-Agents","benchmark_version":"current published task suite","members":[{"configuration":"腾讯官方比较表 · GPT-5.6 Sol","model_id":"m-0510680735d3c79a","normalized_unit_value":"379/10","observation_id":"100759b239352aa8328473780a99c2b5","value":37.9},{"configuration":"腾讯官方比较表 · Hy4 preview","model_id":"m-383f40728e3e5259","normalized_unit_value":"371/10","observation_id":"7cc7a31c8754c107bbefba22dfda40ad","value":37.1},{"configuration":"腾讯官方比较表 · Claude Opus 5","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"209/5","observation_id":"803b3c7ebd5d583be94f718268d22790","value":41.8},{"configuration":"腾讯官方比较表 · Qwen3.8 Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"34","observation_id":"bf4bf643f4907b71184823fd0ebd6b1c","value":34.0},{"configuration":"腾讯官方比较表 · Kimi K3","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"186/5","observation_id":"742baf559c583daecd3dac4f871089d0","value":37.2},{"configuration":"腾讯官方比较表 · GLM-5.3","model_id":"m-c65e3e17492500e9","normalized_unit_value":"381/10","observation_id":"c0d551d2a17a8d88a7eef965f627a78f","value":38.1},{"configuration":"腾讯官方比较表 · Hy3","model_id":"m-d23e83fbb1d3aabc","normalized_unit_value":"122/5","observation_id":"869a19a4a7bb34031859441309a0c7b5","value":24.4},{"configuration":"腾讯官方比较表 · DeepSeek V4 Pro 0813","model_id":"m-f936349cbdb1def9","normalized_unit_value":"162/5","observation_id":"cbcb87856c45df67217ec3b0da377d9c","value":32.4}],"protocol_id":"cc9a245346c3577b6adef0db","unit":"%"},"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66":{"benchmark":"MirrorCode","benchmark_version":"run 0.1.200","members":[{"configuration":"gemini-3.7-flash_high","model_id":"m-16a8dce2a6e7167a","normalized_unit_value":"4111111111111111/10000000000000000","observation_id":"02fcbee985f3da4b53d62ce022f2a290","value":0.4111111111111111},{"configuration":"grok-4.6_xhigh","model_id":"m-3c17c0d3ca7f1abe","normalized_unit_value":"4111111111111111/10000000000000000","observation_id":"0621e61b8874eac3232c0422d1f71677","value":0.4111111111111111},{"configuration":"gemini-4-argon_high","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"5888888888888889/10000000000000000","observation_id":"d2afdbf96101f2fb4499bd4ee1fc2881","value":0.5888888888888889},{"configuration":"grok-4.5_high","model_id":"m-b0c41bb29e51489d","normalized_unit_value":"2013888888888889/6250000000000000","observation_id":"09300d4eee5c7d673f52949c9b3dcb98","value":0.32222222222222224},{"configuration":"gpt-6-luna_max","model_id":"m-e5d076977e951607","normalized_unit_value":"13333333333333333/50000000000000000","observation_id":"9b168b30e93ee9c993180224bf1ba997","value":0.26666666666666666}],"protocol_id":"f59ad8767b1d0f2829ac677e","unit":"比例"},"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e":{"benchmark":"FrontierCode","benchmark_version":"2.0","members":[{"configuration":"Claude Haiku 5.5 (Max)","model_id":"m-016da3c7828e329e","normalized_unit_value":"69865101010101/200000000000000","observation_id":"cbf1fa28fdc821e9836816e2dbc59c73","value":0.349325505050505},{"configuration":"GPT-6.1 Sol (Max)","model_id":"m-0f7beea535abe0d6","normalized_unit_value":"406262878787879/1000000000000000","observation_id":"319a73d6f8fba5d605260942d6b8e601","value":0.406262878787879},{"configuration":"Granite 4.2 3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"0","observation_id":"d294135a5f841df5749d74dee1cdb9e7","value":0},{"configuration":"Granite 4.2 8B","model_id":"m-1d98b8595cf1927d","normalized_unit_value":"0","observation_id":"6a0e7ddb2b9af667874b506e526b6d54","value":0},{"configuration":"GPT-6.1 Terra (Max)","model_id":"m-1dcd13c2fce7d97b","normalized_unit_value":"84640101010101/200000000000000","observation_id":"bf241286cb096d94e2aa92889567c092","value":0.423200505050505},{"configuration":"Muse Wonder Full (Max)","model_id":"m-21b38c9ea818aebe","normalized_unit_value":"377576262626263/1000000000000000","observation_id":"36981a7ad16ac52e29b5d161b5c3a617","value":0.377576262626263},{"configuration":"K2 Horizon 3.7B","model_id":"m-26bf3900b3ceb955","normalized_unit_value":"452739898989899/50000000000000000","observation_id":"908a5049e650c9f2841b87a654313ae7","value":0.00905479797979798},{"configuration":"North Mini Code","model_id":"m-2c518178af26d1d4","normalized_unit_value":"18760101010101/10000000000000000","observation_id":"e2d037eb56a1707481cab5a658eb199f","value":0.0018760101010101},{"configuration":"Inkling Small","model_id":"m-35c4f287acc76c95","normalized_unit_value":"238743686868687/5000000000000000","observation_id":"066fdac9a6f7fc061fce980c5a691276","value":0.0477487373737374},{"configuration":"Nemotron 3.5 Super VL","model_id":"m-36451e3f44503d37","normalized_unit_value":"610914141414141/10000000000000000","observation_id":"35228bba34f7d198ea6ae7b85139d6dc","value":0.0610914141414141},{"configuration":"Gemini 3.8 Flash (High)","model_id":"m-422e5ba87859dd55","normalized_unit_value":"93713888888889/500000000000000","observation_id":"926e007cfcb9b2d7a75e59641d96bd9e","value":0.187427777777778},{"configuration":"GPT-6 Astra (Max)","model_id":"m-5553b7ca5f3e748a","normalized_unit_value":"21187398989899/50000000000000","observation_id":"18ca4c50e781eec01bc5749fbb189315","value":0.42374797979798},{"configuration":"Mistral Medium 3.5","model_id":"m-5c18cc0e1d3391a9","normalized_unit_value":"176893939393939/10000000000000000","observation_id":"c2dd8ca14021780f9edc8dc0975ff8ae","value":0.0176893939393939},{"configuration":"MiMo-V2.6-Flash","model_id":"m-60cd64beda366768","normalized_unit_value":"216152272727273/1000000000000000","observation_id":"d9ee15576eab5d8aeca7fb1035d14154","value":0.216152272727273},{"configuration":"Kimi K3.1 Pro","model_id":"m-66c1aae520b7679a","normalized_unit_value":"63629861111111/250000000000000","observation_id":"2cd7a142e0f06d143bb474c89fce8084","value":0.254519444444444},{"configuration":"Muse Wonder (Max)","model_id":"m-68480fc0423c5034","normalized_unit_value":"27875303030303/100000000000000","observation_id":"3733f77f03c95d14f61b1419adf05cd6","value":0.27875303030303},{"configuration":"Grok 4.7 (Xhigh)","model_id":"m-7187cf2b01d41935","normalized_unit_value":"57315101010101/200000000000000","observation_id":"efab926973de754a7bc887e6a5732e79","value":0.286575505050505},{"configuration":"Mistral Large 4 (High)","model_id":"m-72b806dc67894cf3","normalized_unit_value":"149581313131313/1000000000000000","observation_id":"c316e6a1765e379f023423782e08477b","value":0.149581313131313},{"configuration":"Gemini 4 Argon (High)","model_id":"m-7ff6a9f995dcaa1c","normalized_unit_value":"257076262626263/1000000000000000","observation_id":"aabef6305c7bf9e588cc8a4221b8394d","value":0.257076262626263},{"configuration":"Qwen3.8 Max (0902)","model_id":"m-825009614c016c1d","normalized_unit_value":"234232323232323/1000000000000000","observation_id":"6cbe0b4839fdc145889ebffe2def2a50","value":0.234232323232323},{"configuration":"Nanbeige4.3-33B","model_id":"m-8555d29b166d94e4","normalized_unit_value":"5358851010101/50000000000000","observation_id":"9302ab59d95f9c2a82c7d7c38865f23b","value":0.10717702020202},{"configuration":"Command A+","model_id":"m-8877a9675c79fcad","normalized_unit_value":"0","observation_id":"c8987663cd611e1b09bc6211e110c938","value":0},{"configuration":"MiniMax-M3.1-Flash","model_id":"m-8e76831a5814eec9","normalized_unit_value":"104109090909091/500000000000000","observation_id":"1935649843d5e809cef231b07734be32","value":0.208218181818182},{"configuration":"Claude Fable 5.1 (Max, Default Fallback)","model_id":"m-98ee7005cf45b532","normalized_unit_value":"316248737373737/1000000000000000","observation_id":"468e0160c503684573c12f005950e313","value":0.316248737373737},{"configuration":"Nemotron 3.5 Lightning","model_id":"m-9ff6170077842b96","normalized_unit_value":"0","observation_id":"c503c45f721e54502bd0c3c277c347cb","value":0},{"configuration":"Ling 3.1 Flash 1019","model_id":"m-a3f1f1de993a039e","normalized_unit_value":"170090151515151/1000000000000000","observation_id":"a3fd831de30519528e597eb5e2ceea3e","value":0.170090151515151},{"configuration":"K2 Horizon MoVA 36B A4B","model_id":"m-a67d9ce95a33f246","normalized_unit_value":"112548611111111/2500000000000000","observation_id":"33a3261e51180ab895a82ccd6f7e12a3","value":0.0450194444444444},{"configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","model_id":"m-ad7decc9ae94df12","normalized_unit_value":"13625946969697/1250000000000000","observation_id":"07260be02cf7c09201c1af77d371a386","value":0.0109007575757576},{"configuration":"Inkling (Xhigh)","model_id":"m-b439c804161fac26","normalized_unit_value":"196243686868687/5000000000000000","observation_id":"9b48c9ab902cc3f53ca099a627db7e76","value":0.0392487373737374},{"configuration":"Beam 1.0","model_id":"m-b6a574cc5300d5f1","normalized_unit_value":"553143939393939/10000000000000000","observation_id":"06196596577df812dcdeed161680d16c","value":0.0553143939393939},{"configuration":"Tencent Hy4","model_id":"m-b88efd1e16a0b031","normalized_unit_value":"31447601010101/100000000000000","observation_id":"8ed6fd9401411cecf6b0e548f78e12a5","value":0.31447601010101},{"configuration":"Muse Glimmer (High)","model_id":"m-bf7f21a45c57e0f7","normalized_unit_value":"73715909090909/5000000000000000","observation_id":"eae24b4f084dd40f578e44715aebd484","value":0.0147431818181818},{"configuration":"K-EXAONE 2.0","model_id":"m-c46d8f7f1c3144cb","normalized_unit_value":"207885101010101/25000000000000000","observation_id":"a86b3f52322966e22c9f2052d8fbc893","value":0.00831540404040404},{"configuration":"K2 Horizon 7B","model_id":"m-c58d5bdb4d2e38bc","normalized_unit_value":"4650101010101/400000000000000","observation_id":"d64bab068e454f76d91e1fb18d73dd68","value":0.0116252525252525},{"configuration":"GLM-5.3 (Max)","model_id":"m-c65e3e17492500e9","normalized_unit_value":"5396773989899/25000000000000","observation_id":"326d68113ecd602bd7d85d3d2b454d3f","value":0.21587095959596},{"configuration":"Qwen3.8 27B (Xhigh)","model_id":"m-c947385433db3f03","normalized_unit_value":"149142929292929/1000000000000000","observation_id":"328d104cb62d50eef3ae2e92cc4db522","value":0.149142929292929},{"configuration":"DeepSeek V4.1 Flash 1017 (Max)","model_id":"m-caf1ef1f3bdf5452","normalized_unit_value":"17794476010101/62500000000000","observation_id":"981d7322c13b53f8d5edfa1fdb6e95e5","value":0.284711616161616},{"configuration":"GLM-5.3-Flash","model_id":"m-cf177eb625d4bd3e","normalized_unit_value":"35214090909091/200000000000000","observation_id":"32536ff0b3b8576673f9f05323026239","value":0.176070454545455},{"configuration":"Claude Opus 5.5 (Max, Default Fallback)","model_id":"m-cfe0652e9d685783","normalized_unit_value":"116089898989899/250000000000000","observation_id":"a1969d9eb66e2bed1a98e36808c17360","value":0.464359595959596},{"configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","model_id":"m-d123610543e5f5ed","normalized_unit_value":"208496717171717/500000000000000","observation_id":"2bdac0e90d63fa507e78af95b1c832a5","value":0.416993434343434},{"configuration":"K2 Horizon 375B A23B","model_id":"m-d2febb93118535a2","normalized_unit_value":"23694292929293/200000000000000","observation_id":"b9e8a0463a865f6221caaaf1d3aecc9d","value":0.118471464646465},{"configuration":"MiniMax-M3","model_id":"m-da421acad97d607b","normalized_unit_value":"219031313131313/1000000000000000","observation_id":"3f103c270013e8e3a189c86aaaac75c3","value":0.219031313131313},{"configuration":"MiMo-V2.6-Pro","model_id":"m-dabee5e3d87d8a7a","normalized_unit_value":"239701262626263/1000000000000000","observation_id":"f7aceff949a2ebec04ba4fe1d1d7e419","value":0.239701262626263},{"configuration":"Muse Spark Open (Xhigh)","model_id":"m-db208dec31031420","normalized_unit_value":"760421717171717/10000000000000000","observation_id":"ca9b3bebd5cf2b1f50a3f98180ff0076","value":0.0760421717171717},{"configuration":"Gemini 4 Carbon (High)","model_id":"m-dea86e466cdc70f2","normalized_unit_value":"405606060606061/1000000000000000","observation_id":"54642e364cbdcff99c9cf0483b73a212","value":0.405606060606061},{"configuration":"GPT-6 Luna (Max)","model_id":"m-e5d076977e951607","normalized_unit_value":"251747727272727/1000000000000000","observation_id":"91532a2952c890cc776404a1ca00195c","value":0.251747727272727},{"configuration":"Gemini 3.5 Flash-Lite","model_id":"m-e7f69d0dac9fe632","normalized_unit_value":"602982323232323/10000000000000000","observation_id":"26e16bd219d1e573463da6c55739a16d","value":0.0602982323232323},{"configuration":"Inkling 2 Small","model_id":"m-f33f6c88c27b2b9a","normalized_unit_value":"25352601010101/100000000000000","observation_id":"0fe380f78c53ebf4d268a83df481e0ad","value":0.25352601010101},{"configuration":"Step 5 Preview","model_id":"m-f892623396612d9b","normalized_unit_value":"120010101010101/1000000000000000","observation_id":"4086551829a96ee8ea164e44d1e7c1c5","value":0.120010101010101},{"configuration":"DeepSeek V4 Pro 0813 (Max)","model_id":"m-f936349cbdb1def9","normalized_unit_value":"170742929292929/1000000000000000","observation_id":"2e337d0b2b65e72e819796e03cdc53e7","value":0.170742929292929}],"protocol_id":"7455c4b3b3db4769bfc4048f","unit":"比例"},"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647":{"benchmark":"MMLU-Pro","benchmark_version":"current maintainer dataset","members":[{"configuration":"Qwen3.7-Plus","model_id":"m-7190ddc26a5a37a0","normalized_unit_value":"177/2","observation_id":"7be3425578af009ce953f32bc09ba9aa","value":88.5},{"configuration":"Qwen3.7-Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"448/5","observation_id":"3e7111ab4e0b4463d98c531d78381779","value":89.6}],"protocol_id":"def1c3d4f8bd615659844ee7","unit":"%"},"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b":{"benchmark":"IFEval","benchmark_version":"current maintainer dataset","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"937/10","observation_id":"55369a687291e4012c741d522a892ee5","value":93.7},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"454/5","observation_id":"0c588d45b55a074bc2915de39e2cf5ff","value":90.8},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"867/10","observation_id":"c5c1fcc77312c8bfdb8f9a2733828c50","value":86.7},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"467/5","observation_id":"a24e132d4558f677a1ce0a399471dbb9","value":93.4}],"protocol_id":"b7479430c87ac4524f0a5b12","unit":"%"},"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8":{"benchmark":"MCP Atlas","benchmark_version":"1000 tasks","members":[{"configuration":"GPT-5.6 Sol (max)","model_id":"m-0510680735d3c79a","normalized_unit_value":"418/5","observation_id":"c608d42d9d5288f84ff9a46cf0c9f453","value":83.6},{"configuration":"GLM-5.2 (max)","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"413/5","observation_id":"d57170302e825b662bfe4acca3fbd398","value":82.6},{"configuration":"Claude Fable 5 (max, w/ fallback)","model_id":"m-86c0c8bf10aa4a0d","normalized_unit_value":"847/10","observation_id":"fdab9cc10508c8861689ba5d61adcb18","value":84.7},{"configuration":"Kimi K3 (max)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"421/5","observation_id":"0949556ec5085216686632a53c4b2aa4","value":84.2},{"configuration":"Solar Open 2 250B-A15B","model_id":"m-b7faaf31060e95f7","normalized_unit_value":"291/5","observation_id":"b96a70ae7b2ded290b5f169ceedf1e03","value":58.2},{"configuration":"Claude Opus 4.8 (max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"418/5","observation_id":"2ba9fb0fcdacf536fe9c61d211eab736","value":83.6}],"protocol_id":"f6ed528afdd4dc3202de03d2","unit":"%"},"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489":{"benchmark":"MATH-500","benchmark_version":"500-item subset","members":[{"configuration":"granite-4.2-3B","model_id":"m-12d2cbffeaae72d5","normalized_unit_value":"97","observation_id":"48e7cd14f60fe5daf4244ca5ca1c19b2","value":97.0},{"configuration":"LFM2.5-8B-A1B","model_id":"m-161fc8601724c921","normalized_unit_value":"466/5","observation_id":"158a7a0bcbb4fdee990dff1e880fb156","value":93.2},{"configuration":"MiniCPM5-2B","model_id":"m-407d9206347b790d","normalized_unit_value":"473/5","observation_id":"1f0d5271a4aa3a380367a096ae7ae210","value":94.6},{"configuration":"LFM2.5-2.6B","model_id":"m-d0874ed91fd69e8e","normalized_unit_value":"448/5","observation_id":"d2a6fbf18a7d28ad26aab8f6363bd884","value":89.6}],"protocol_id":"a181f644118455cd59af445a","unit":"%"},"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833":{"benchmark":"AA-Briefcase","benchmark_version":"1.1","members":[{"configuration":"GPT-5.6 Sol (Medium)","model_id":"m-0510680735d3c79a","normalized_unit_value":"1590909090909091/5000000000000000","observation_id":"02e903393ff93c77cfd706e931424b8b","value":0.3181818181818182},{"configuration":"Gemini 3.7 Flash (High)","model_id":"m-16a8dce2a6e7167a","normalized_unit_value":"7272727272727273/20000000000000000","observation_id":"d43d801b58cb460394b97c15079fc130","value":0.36363636363636365},{"configuration":"Claude Sonnet 5 (Xhigh)","model_id":"m-2cd37409a483435a","normalized_unit_value":"2/5","observation_id":"14b2d91b066241cbb4ba0914b53ae0a4","value":0.4},{"configuration":"Muse Spark 1.2 (Xhigh)","model_id":"m-2f22b07da3af863c","normalized_unit_value":"22575757575757577/50000000000000000","observation_id":"2a4003430f96eaca6cf7fbdac62ade55","value":0.45151515151515154},{"configuration":"DeepSeek V4 Flash 0731 (Max)","model_id":"m-3535a3e7891af5ef","normalized_unit_value":"10378787878787879/25000000000000000","observation_id":"239fbb8019bce72c55d5b79e4b9186e5","value":0.41515151515151516},{"configuration":"Grok 4.6 (Low)","model_id":"m-3c17c0d3ca7f1abe","normalized_unit_value":"201010101010101/500000000000000","observation_id":"45f7b80c7cc89c267cb8e0d74593b6e4","value":0.402020202020202},{"configuration":"Claude Opus 5 (Medium)","model_id":"m-568e0eb3a95b5113","normalized_unit_value":"24242424242424243/50000000000000000","observation_id":"a01ae7f3c128573c113df7e946ef1073","value":0.48484848484848486},{"configuration":"GLM-5.2 (Max)","model_id":"m-5cf6d704eda6778e","normalized_unit_value":"1803030303030303/5000000000000000","observation_id":"5e2ad6a27f43522ccb21c4f280e4613e","value":0.3606060606060606},{"configuration":"GPT-5.6 Luna (Max)","model_id":"m-65a8fc1dec61133f","normalized_unit_value":"1898989898989899/5000000000000000","observation_id":"06516697c2ae14f64607af00c620c8a8","value":0.3797979797979798},{"configuration":"Gemini 3.6 Flash (High)","model_id":"m-77b5dbc317075e9d","normalized_unit_value":"351010101010101/1250000000000000","observation_id":"ec2017cdf9726d6330e3b72963cbf545","value":0.2808080808080808},{"configuration":"Gemini 3.5 Flash (Medium)","model_id":"m-8d99798a7e24690c","normalized_unit_value":"27635619242579323/100000000000000000","observation_id":"a4acfc86f781d098a642bc8bcfd57f42","value":0.27635619242579323},{"configuration":"DeepSeek V4.1 Flash (Max)","model_id":"m-99183a5b3b302b0d","normalized_unit_value":"9038854805725971/20000000000000000","observation_id":"59421edaa49a1a1c8543f29553d821c6","value":0.45194274028629855},{"configuration":"Qwen3.8 Max","model_id":"m-99ca714b8835cfb9","normalized_unit_value":"49595959595959593/100000000000000000","observation_id":"0d1d30147df15075a1ed15467a0ef1b4","value":0.49595959595959593},{"configuration":"Kimi K3 (Low)","model_id":"m-9e460c49820cbe4e","normalized_unit_value":"1303030303030303/5000000000000000","observation_id":"04f715495ca6c55181e1b42a64312d78","value":0.2606060606060606},{"configuration":"Qwen3.7 Max","model_id":"m-a3a34c5608c68d32","normalized_unit_value":"11868686868686869/50000000000000000","observation_id":"680700051014bfbbbe521e958802f31c","value":0.23737373737373738},{"configuration":"Grok 4.5 (High)","model_id":"m-b0c41bb29e51489d","normalized_unit_value":"4070707070707071/10000000000000000","observation_id":"99d362c58056009c0378c89776bd8bfa","value":0.4070707070707071},{"configuration":"Muse Spark 1.3 (Xhigh)","model_id":"m-c0580b01221ed260","normalized_unit_value":"2580808080808081/5000000000000000","observation_id":"0619cb6c1df6def76a850c294e50de63","value":0.5161616161616162},{"configuration":"Claude Opus 4.8 (Max)","model_id":"m-cccbc1ce3042d216","normalized_unit_value":"19343434343434343/50000000000000000","observation_id":"52ae106aa8230eabcaac4637b6d547ea","value":0.38686868686868686},{"configuration":"GPT-5.6 Terra (Low)","model_id":"m-f8dbf8d16fd37aac","normalized_unit_value":"11666666666666667/50000000000000000","observation_id":"56c8ab5d498b1025e9684eb22709b6c7","value":0.23333333333333334}],"protocol_id":"66efeab01f142e89eb5e2362","unit":"比例"}},"reported_numeric_observation_count":1272,"reviewed_at":"2026-10-29T10:00:59.694526+00:00","revision":"2e0764d7fadf16f8cafb703aac0d0e8885987809201805ce4874b3f68bd79cec","rows":[{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":153,"numerator":15100},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench-Science","row_ids":["416c079f92e7711d76f35d3a6da3f408"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"HLE","row_ids":["2541b171f3a9b70175c5eff4ce5a2a47"],"score":100.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4900},"family":"CritPt","row_ids":["29eed13fef0ca4e5917825122c33b6b0"],"score":96.08}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","peer_count":64,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","peer_count":52,"percentile":96.08,"percentile_exact":{"denominator":51,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69}],"score":98.69,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":3564,"numerator":321485},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7400},"family":"GDPval-AA","row_ids":["5c7968ea3f6404bb3d978285ac42aa05"],"score":91.36},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":375},"family":"Terminal-Bench","row_ids":["2db4ed4fb97d84958b935e83df5db43d"],"score":93.75},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["1b1f1f3e062850814bcd282684d4e879"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":1000},"family":"AA-Briefcase","row_ids":["fe408037e14234a002d5a2d7a3b186f0"],"score":90.91},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":75},"family":"OSWorld","row_ids":["f7b8bce78b6bbb414e44451ecece6456"],"score":75.0}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","peer_count":82,"percentile":91.36,"percentile_exact":{"denominator":81,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","peer_count":49,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":78,"percentile":90.91,"percentile_exact":{"denominator":11,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5}],"score":90.2,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":4400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4400},"family":"FrontierCode","row_ids":["54642e364cbdcff99c9cf0483b73a212"],"score":89.8}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","peer_count":50,"percentile":89.8,"percentile_exact":{"denominator":49,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061}],"score":89.8,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":94350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1750},"family":"GDP.pdf","row_ids":["2ef7660125e00e7ecf9ebbabeb38c644"],"score":92.11},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":5050},"family":"AA-Omniscience","row_ids":["00d59d147ff0b7cd3c12bdcaa9a4b516","91da142d2c0422108afa80f33bfb9bb4"],"score":95.28}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339}],"score":93.69,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":100},"family":"MMMU-Pro","row_ids":["34112c44d835a7ef36f154c3c42aa2b1"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["98879b4fa09dc7b005731381bea066e0"],"score":100.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5}],"score":100.0,"weight":10}],"eligible":true,"id":"m-dea86e466cdc70f2","in_window":true,"name":"Gemini 4 Carbon","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":1,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"416c079f92e7711d76f35d3a6da3f408","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"0.1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.647619047619048,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-4-carbon:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.647619047619048},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2541b171f3a9b70175c5eff4ce5a2a47","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"Diamond","peer_count":64,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.691,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-carbon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29eed13fef0ca4e5917825122c33b6b0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2 Hard","peer_count":52,"percentile":96.08,"percentile_exact":{"denominator":51,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.69,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-carbon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.69},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c7968ea3f6404bb3d978285ac42aa05","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v2.1","peer_count":82,"percentile":91.36,"percentile_exact":{"denominator":81,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1732.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-carbon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1732.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 4 Carbon / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2db4ed4fb97d84958b935e83df5db43d","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"4.0","peer_count":49,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":63.13,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"63.13%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":63.13},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b1f1f3e062850814bcd282684d4e879","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7781660572624064,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-carbon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7781660572624064},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe408037e14234a002d5a2d7a3b186f0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":78,"percentile":90.91,"percentile_exact":{"denominator":11,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1610.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-carbon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1610.63},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f7b8bce78b6bbb414e44451ecece6456","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.1","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":81.5,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"81.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":81.5},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"54642e364cbdcff99c9cf0483b73a212","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"2.0","peer_count":50,"percentile":89.8,"percentile_exact":{"denominator":49,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.405606060606061,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-carbon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.405606060606061},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ef7660125e00e7ecf9ebbabeb38c644","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current AA protocol","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.284,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-carbon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"00d59d147ff0b7cd3c12bdcaa9a4b516","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5826666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5826666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Carbon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"91da142d2c0422108afa80f33bfb9bb4","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8857827476038339,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-carbon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-carbon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8857827476038339},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Carbon (Medium)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"34112c44d835a7ef36f154c3c42aa2b1","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current Standard 10","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.892485549132948,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.892485549132948},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Carbon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"98879b4fa09dc7b005731381bea066e0","included":true,"model_id":"m-dea86e466cdc70f2","original_version_label":"current published dataset","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":72.5,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"72.5%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Carbon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Carbon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":72.5}],"release_date":"2026-10-28","release_status":"published","release_verified":true,"result_updated_at":"2026-10-28","score":93.92,"score_exact":{"denominator":11958374736,"numerator":1123155312275},"source_id":"published/aa:gemini-4-carbon"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":61047,"numerator":5798875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1775},"family":"Terminal-Bench-Science","row_ids":["21af8993ff5f3e38ecc3caf66271b208"],"score":93.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":6200},"family":"HLE","row_ids":["3522855b1814574fd5b3733b32b77769"],"score":98.41},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4750},"family":"CritPt","row_ids":["46e4906d019b793c19b1d443cd35b178"],"score":93.14}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","peer_count":64,"percentile":98.41,"percentile_exact":{"denominator":63,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68}],"score":94.99,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":302841,"numerator":28338056},"families":[{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":100},"family":"GDPval-AA","row_ids":["b44123b055ba24f02daafca81efcb87b"],"score":100.0},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":96},"family":"Terminal-Bench","row_ids":["322131c4aa07af285df46f472061072d"],"score":96.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1750},"family":"AutomationBench","row_ids":["f100a726506576804779fdbd76208d6e"],"score":92.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7600},"family":"AA-Briefcase","row_ids":["5c16c97705a8f70907dc1ba381949448"],"score":98.7},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":250},"family":"OSWorld","row_ids":["cb71933d15a49daa0088c6b64854af7f"],"score":83.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2100},"family":"AA-AnalystAgent","row_ids":["1a02ab6144866fd1b960f19d8fd21513"],"score":91.3}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","peer_count":82,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"4.0","peer_count":76,"percentile":96.0,"percentile_exact":{"denominator":1,"numerator":96},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":78,"percentile":98.7,"percentile_exact":{"denominator":77,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.1","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","peer_count":24,"percentile":91.3,"percentile_exact":{"denominator":23,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625}],"score":93.57,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":1275},"families":[{"benchmark_version":"2.0","exact":{"denominator":1,"numerator":100},"family":"FrontierCode","row_ids":["a1969d9eb66e2bed1a98e36808c17360"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1250},"family":"Vals Code Migration","row_ids":["d2e15e2796cf44b3aabbae2da4bd5b3f"],"score":96.15}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.0","peer_count":50,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"1.1","peer_count":27,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01}],"score":98.08,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2822875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3425},"family":"GDP.pdf","row_ids":["77cd16d89e49025a3285c182b2da553d"],"score":90.13},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1550},"family":"SimpleQA Verified","row_ids":["144b27b350bf50d4c19f78baa94c183a"],"score":91.18},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3500},"family":"AA-Omniscience","row_ids":["3247fbdae3f488e2395593b52951cc4b","035a4d676bedc8d21713e0b6aa5bc0ad"],"score":66.04}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227}],"score":82.45,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":459,"numerator":37630},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"Humanity Sixth Sense","row_ids":["78db71948ffd6a9f7a24a59779294d90"],"score":88.89},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1650},"family":"MMMU-Pro","row_ids":["3dabe34a660554a2d248f440d9ae144c"],"score":97.06},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":60},"family":"Chartography","row_ids":["d2e4430a89fa4e3972bc81f62bbdad07"],"score":60.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","peer_count":28,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":60.0,"percentile_exact":{"denominator":1,"numerator":60},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4}],"score":81.98,"weight":10}],"eligible":true,"id":"m-cfe0652e9d685783","in_window":true,"name":"Claude Opus 5.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":2,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"21af8993ff5f3e38ecc3caf66271b208","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row3","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3522855b1814574fd5b3733b32b77769","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"Diamond","peer_count":64,"percentile":98.41,"percentile_exact":{"denominator":63,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.679,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-opus-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.679},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"46e4906d019b793c19b1d443cd35b178","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-opus-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b44123b055ba24f02daafca81efcb87b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v2.1","peer_count":82,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1866.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-opus-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1866.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"322131c4aa07af285df46f472061072d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"4.0","peer_count":76,"percentile":96.0,"percentile_exact":{"denominator":1,"numerator":96},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.595959595959596,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-opus-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.595959595959596},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f100a726506576804779fdbd76208d6e","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":92.11,"percentile_exact":{"denominator":19,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6953868128636069,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-opus-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6953868128636069},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c16c97705a8f70907dc1ba381949448","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":78,"percentile":98.7,"percentile_exact":{"denominator":77,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1799.89,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-opus-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1799.89},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cb71933d15a49daa0088c6b64854af7f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.1","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":81.8,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"81.8% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":81.8},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1a02ab6144866fd1b960f19d8fd21513","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA pass^5","peer_count":24,"percentile":91.3,"percentile_exact":{"denominator":23,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-opus-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a1969d9eb66e2bed1a98e36808c17360","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"2.0","peer_count":50,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.464359595959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-opus-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.464359595959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Opus 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"d2e15e2796cf44b3aabbae2da4bd5b3f","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"1.1","peer_count":27,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":81.01,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"81.01%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row2","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":81.01},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"77cd16d89e49025a3285c182b2da553d","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-opus-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"144b27b350bf50d4c19f78baa94c183a","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"run 1.2.0","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.722,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T18:27:39.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0de4d0fc7db2d9072fdcb34e"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0de4d0fc7db2d9072fdcb34e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.722},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3247fbdae3f488e2395593b52951cc4b","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":98.11,"percentile_exact":{"denominator":53,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.655,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.655},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"035a4d676bedc8d21713e0b6aa5bc0ad","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4966183574879227,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-opus-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4966183574879227},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5.5 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"78db71948ffd6a9f7a24a59779294d90","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current public protocol","peer_count":28,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":44.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cc6d207297718c264ee22901"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cc6d207297718c264ee22901","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":44.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Opus 5.5 (Max, Default Fallback)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"3dabe34a660554a2d248f440d9ae144c","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"current Standard 10","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.876878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-opus-5-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"claude-opus-5-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.876878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Opus 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"d2e4430a89fa4e3972bc81f62bbdad07","included":true,"model_id":"m-cfe0652e9d685783","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":60.0,"percentile_exact":{"denominator":1,"numerator":60},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":64.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"64.4% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column3","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":64.4}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":"2026-09-28","score":91.44,"score_exact":{"denominator":42566119596,"numerator":3892435594141},"source_id":"anthropic/claude-opus-5-5"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":3958475},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":100},"family":"FrontierMath","row_ids":["cacd9412ff6ba035ea9e826693b09343"],"score":100.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1850},"family":"Terminal-Bench-Science","row_ids":["77175d98c0074eb53a919fa599931946"],"score":97.37},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5900},"family":"HLE","row_ids":["9165499121e5e198448d4536365d9c16"],"score":93.65},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":5000},"family":"CritPt","row_ids":["96d62eb6c5d675afc73f8103f6dbc627"],"score":98.04}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","peer_count":39,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","peer_count":64,"percentile":93.65,"percentile_exact":{"denominator":63,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","peer_count":52,"percentile":98.04,"percentile_exact":{"denominator":51,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695}],"score":97.26,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4542615,"numerator":363697727},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2000},"family":"GDPval-AA","row_ids":["eaa5c4189ba4b51426be70de02910e04"],"score":74.07},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":272},"family":"Terminal-Bench","row_ids":["d04f5531bb77f209689af5059ec9471d"],"score":90.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1575},"family":"AutomationBench","row_ids":["4bd5c8368df638bd0c064dd06c306d48"],"score":82.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6400},"family":"AA-Briefcase","row_ids":["c7f025daa6b0e96ab52dc7df7e3bb96a"],"score":83.12},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1600},"family":"AA-AnalystAgent","row_ids":["9b9634fd2b2367e076a491de1e4f02e6"],"score":69.57}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","peer_count":82,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","peer_count":76,"percentile":90.67,"percentile_exact":{"denominator":3,"numerator":272},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":78,"percentile":83.12,"percentile_exact":{"denominator":77,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","peer_count":24,"percentile":69.57,"percentile_exact":{"denominator":23,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5}],"score":80.06,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":57425},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4500},"family":"FrontierCode","row_ids":["319a73d6f8fba5d605260942d6b8e601"],"score":91.84},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1150},"family":"Vals Code Migration","row_ids":["e330fa37e52713545b5fec3f612240b8"],"score":88.46}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","peer_count":50,"percentile":91.84,"percentile_exact":{"denominator":49,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","peer_count":27,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32}],"score":90.15,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":12153775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3675},"family":"GDP.pdf","row_ids":["f0c0c3073a678bbcb3f5f855da9adff8"],"score":96.71},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1600},"family":"SimpleQA Verified","row_ids":["509c52385e7c5559b4414a3e4bdb93cc"],"score":94.12},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"MultiChallenge","row_ids":["2a39ee205e56064a7ebeddb7db41c3ee"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3400},"family":"AA-Omniscience","row_ids":["ad20786a83f6b0b25b2f134cb2f6c095","8c01a1619e7c484574ab42025543f92a"],"score":64.15}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075}],"score":88.74,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":42175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2500},"family":"Humanity Sixth Sense","row_ids":["3a0983451c53206a0fa5061a863b8a25"],"score":92.59},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1550},"family":"MMMU-Pro","row_ids":["9a17a04ad8c7b9671f59a85153438a32"],"score":91.18}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","peer_count":28,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335}],"score":91.88,"weight":10}],"eligible":true,"id":"m-0f7beea535abe0d6","in_window":true,"name":"GPT-6.1 Sol","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-29","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":3,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cacd9412ff6ba035ea9e826693b09343","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":1.0,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f0855468c90c22d36e2fa262"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"f0855468c90c22d36e2fa262","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":1.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"77175d98c0074eb53a919fa599931946","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"0.1.0","peer_count":39,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":52.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"52.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row2"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row2","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":52.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9165499121e5e198448d4536365d9c16","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"Diamond","peer_count":64,"percentile":93.65,"percentile_exact":{"denominator":63,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.588,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-1-sol:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.588},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"96d62eb6c5d675afc73f8103f6dbc627","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2 Hard","peer_count":52,"percentile":98.04,"percentile_exact":{"denominator":51,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.695,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-sol:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.695},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eaa5c4189ba4b51426be70de02910e04","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v2.1","peer_count":82,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1608.89,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-sol:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1608.89},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d04f5531bb77f209689af5059ec9471d","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"4.0","peer_count":76,"percentile":90.67,"percentile_exact":{"denominator":3,"numerator":272},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.560606060606061,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-sol:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.560606060606061},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4bd5c8368df638bd0c064dd06c306d48","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6486844882398171,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-sol:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6486844882398171},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c7f025daa6b0e96ab52dc7df7e3bb96a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":78,"percentile":83.12,"percentile_exact":{"denominator":77,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1555.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-sol:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1555.09},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9b9634fd2b2367e076a491de1e4f02e6","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA pass^5","peer_count":24,"percentile":69.57,"percentile_exact":{"denominator":23,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-1-sol:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"319a73d6f8fba5d605260942d6b8e601","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"2.0","peer_count":50,"percentile":91.84,"percentile_exact":{"denominator":49,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.406262878787879,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-sol:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.406262878787879},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6.1 Sol / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"e330fa37e52713545b5fec3f612240b8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"1.1","peer_count":27,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":77.32,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"77.32%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row4","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":77.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f0c0c3073a678bbcb3f5f855da9adff8","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-sol:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"509c52385e7c5559b4414a3e4bdb93cc","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"run 1.2.0","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.739,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T22:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9bb6c3f1cced38ca5850dc41"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9bb6c3f1cced38ca5850dc41","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.739},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"gpt 6.1 Sol","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2a39ee205e56064a7ebeddb7db41c3ee","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current LLM protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":81.95,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"260b9050423d667fbaf49f21"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"260b9050423d667fbaf49f21","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":81.95},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ad20786a83f6b0b25b2f134cb2f6c095","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6183333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6183333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Sol (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8c01a1619e7c484574ab42025543f92a","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48122270742358075,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-sol:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48122270742358075},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6.1-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"3a0983451c53206a0fa5061a863b8a25","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current public protocol","peer_count":28,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":46.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e6f336acf8f3983632b182e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0e6f336acf8f3983632b182e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":46.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Sol (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9a17a04ad8c7b9671f59a85153438a32","included":true,"model_id":"m-0f7beea535abe0d6","original_version_label":"current Standard 10","peer_count":35,"percentile":91.18,"percentile_exact":{"denominator":17,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.859537572254335,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-sol:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-sol:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.859537572254335}],"release_date":"2026-09-29","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":89.3,"score_exact":{"denominator":496604728620,"numerator":44346438312619},"source_id":"openai/gpt-6.1-sol"},{"benchmark_count":21,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":100},"family":"FrontierMath","row_ids":["0f175b01de79f7689d7f750dac02d89b"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"GPQA Diamond","row_ids":["c863d641570942861efdba9c19093efa"],"score":100.0},{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench-Science","row_ids":["33b37c700e39b0d0c6a88f59bf14c960"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":100},"family":"HLE","row_ids":["955cc9868301478204da3b4815a3ad3c"],"score":100.0},{"benchmark_version":"2 Hard","exact":{"denominator":1,"numerator":100},"family":"CritPt","row_ids":["0a974b1036bafb40411da54c5245579a"],"score":100.0}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","peer_count":39,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","peer_count":52,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705}],"score":100.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":248026779,"numerator":19458728500},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1100},"family":"tau3-Banking","row_ids":["388b55b5431573f02c7a1415569151aa"],"score":84.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5800},"family":"GDPval-AA","row_ids":["b076d23f5b3bf4c32b53727e95233455"],"score":71.6},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":275},"family":"Terminal-Bench","row_ids":["28b064c611914ab9c6c8b9c66db69c4f"],"score":91.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1725},"family":"AutomationBench","row_ids":["fb60ba99063f49f876c4c8d56120f0f3"],"score":90.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6500},"family":"AA-Briefcase","row_ids":["e3f8cbf64c84e82bb4a86c035247a89a"],"score":84.42},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":50},"family":"OSWorld","row_ids":["2ed0ffc858d7ba150a21822c243ebff1"],"score":50.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1750},"family":"AA-AnalystAgent","row_ids":["15975e3d42ab5d82a67bcbe277dd1b83"],"score":76.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","peer_count":82,"percentile":71.6,"percentile_exact":{"denominator":81,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","peer_count":49,"percentile":91.67,"percentile_exact":{"denominator":3,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":90.79,"percentile_exact":{"denominator":19,"numerator":1725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":78,"percentile":84.42,"percentile_exact":{"denominator":77,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125}],"score":78.45,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":1911,"numerator":184900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4800},"family":"FrontierCode","row_ids":["18ca4c50e781eec01bc5749fbb189315"],"score":97.96},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1200},"family":"Vals Code Migration","row_ids":["bc107d6862df5282c7cf8b968206739b"],"score":92.31},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["948d62640ac1b6f74166a175a86cf6a7"],"score":100.0}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","peer_count":50,"percentile":97.96,"percentile_exact":{"denominator":49,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":27,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":9,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1}],"score":96.76,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2954325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3675},"family":"GDP.pdf","row_ids":["96f5c4d98f214c070d53868688e0a1a3"],"score":96.71},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1650},"family":"SimpleQA Verified","row_ids":["3557cefe053ef3a2ca1c95595065295f"],"score":97.06},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3450},"family":"AA-Omniscience","row_ids":["eec9e765997481a4684eb7ff4710087d","6e1af818db0c40a97c73d45f3dd92647"],"score":65.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958}],"score":86.29,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":1377,"numerator":110350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2600},"family":"Humanity Sixth Sense","row_ids":["35c55ac70a4db5208200b7b35fe696c9"],"score":96.3},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1600},"family":"MMMU-Pro","row_ids":["ec04ac79e57d7b9ceb9bc5c32ee3fd30"],"score":94.12},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"Chartography","row_ids":["600568174d3d7470e372c6b4483b5633"],"score":50.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","peer_count":28,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0}],"score":80.14,"weight":10}],"eligible":true,"id":"m-5553b7ca5f3e748a","in_window":true,"name":"GPT-6 Astra","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":4,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"reasoning","direction":"higher_better","id":"0f175b01de79f7689d7f750dac02d89b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":97.6,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":97.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c863d641570942861efdba9c19093efa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.960606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-6-astra:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.960606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"33b37c700e39b0d0c6a88f59bf14c960","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"0.1.0","peer_count":39,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":62.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"62.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row1","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":62.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-astra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"955cc9868301478204da3b4815a3ad3c","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"Diamond","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":60.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"906936af39b00fbe77f40c81"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"906936af39b00fbe77f40c81","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":60.6},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0a974b1036bafb40411da54c5245579a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2 Hard","peer_count":52,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.705,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-astra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.705},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"388b55b5431573f02c7a1415569151aa","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.414432989690722,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-6-astra:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.414432989690722},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b076d23f5b3bf4c32b53727e95233455","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v2.1","peer_count":82,"percentile":71.6,"percentile_exact":{"denominator":81,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1583.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-astra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1583.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Astra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28b064c611914ab9c6c8b9c66db69c4f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"4.0","peer_count":49,"percentile":91.67,"percentile_exact":{"denominator":3,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":59.6,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"59.60%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":59.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb60ba99063f49f876c4c8d56120f0f3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":90.79,"percentile_exact":{"denominator":19,"numerator":1725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6849174806688876,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-astra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6849174806688876},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3f8cbf64c84e82bb4a86c035247a89a","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":78,"percentile":84.42,"percentile_exact":{"denominator":77,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1568.59,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-astra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1568.59},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"2ed0ffc858d7ba150a21822c243ebff1","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.1","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":80.9,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"80.9%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":80.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"15975e3d42ab5d82a67bcbe277dd1b83","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-astra:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"18ca4c50e781eec01bc5749fbb189315","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"2.0","peer_count":50,"percentile":97.96,"percentile_exact":{"denominator":49,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.42374797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-astra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42374797979798},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Astra / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"bc107d6862df5282c7cf8b968206739b","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":27,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":80.58,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"80.58%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row3"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row3","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":80.58},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑6 Astra","dimension":"coding","direction":"higher_better","id":"948d62640ac1b6f74166a175a86cf6a7","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"1.1","peer_count":9,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":74.1,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column1","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":74.1},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96f5c4d98f214c070d53868688e0a1a3","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current AA protocol","peer_count":77,"percentile":96.71,"percentile_exact":{"denominator":38,"numerator":3675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.31,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-astra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-astra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3557cefe053ef3a2ca1c95595065295f","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"run 1.2.0","peer_count":35,"percentile":97.06,"percentile_exact":{"denominator":17,"numerator":1650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.756,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-30","result_timestamp":"2026-08-30T14:57:43.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9e14e902ceb6bf6bb573f557"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9e14e902ceb6bf6bb573f557","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.756},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"eec9e765997481a4684eb7ff4710087d","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6156666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6156666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Astra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e1af818db0c40a97c73d45f3dd92647","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5021682567215958,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-astra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5021682567215958},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Astra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"35c55ac70a4db5208200b7b35fe696c9","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current public protocol","peer_count":28,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":53.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"5105cb966b7235e54a16a771"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"5105cb966b7235e54a16a771","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":53.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Astra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ec04ac79e57d7b9ceb9bc5c32ee3fd30","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current Standard 10","peer_count":35,"percentile":94.12,"percentile_exact":{"denominator":17,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.86878612716763,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-astra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-astra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.86878612716763},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"GPT-6 Astra / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"600568174d3d7470e372c6b4483b5633","included":true,"model_id":"m-5553b7ca5f3e748a","original_version_label":"current published dataset","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.0,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.0%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :GPT-6 Astra"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :GPT-6 Astra","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.0}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":"2026-10-28","score":89.24,"score_exact":{"denominator":21283059798,"numerator":1899215940095},"source_id":"openai/gpt-6-astra"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":5967,"numerator":520600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["0caa25aeefa10afc49082f57afb1971b"],"score":84.62},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":800},"family":"HLE","row_ids":["4e26762c5adc39cb8336491e6db427a8"],"score":88.89},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1500},"family":"CritPt","row_ids":["f2c7d3b6f687d90b18b5ecae64664442"],"score":88.24}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","peer_count":64,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","peer_count":52,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565}],"score":87.25,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3792096,"numerator":357306625},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7900},"family":"GDPval-AA","row_ids":["fa97f8f99150ab638ce3d789e781973c"],"score":97.53},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":2125},"family":"Terminal-Bench","row_ids":["07f1cdf7bbcb0da17ca1fb2ddf289fbe"],"score":88.54},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1775},"family":"AutomationBench","row_ids":["9d53e1ef4d2e9c319b7f0dacff98d83c"],"score":93.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7500},"family":"AA-Briefcase","row_ids":["4f4f6c3a22e75b41412d7850bc4b9cf5"],"score":97.4}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","peer_count":82,"percentile":97.53,"percentile_exact":{"denominator":81,"numerator":7900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","peer_count":49,"percentile":88.54,"percentile_exact":{"denominator":24,"numerator":2125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":78,"percentile":97.4,"percentile_exact":{"denominator":77,"numerator":7500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32}],"score":94.22,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":637,"numerator":51225},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4300},"family":"FrontierCode","row_ids":["36981a7ad16ac52e29b5d161b5c3a617"],"score":87.76},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":950},"family":"Vals Code Migration","row_ids":["3d507b92bee70e08544b9eb65b20267a"],"score":73.08}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","peer_count":50,"percentile":87.76,"percentile_exact":{"denominator":49,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","peer_count":27,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79}],"score":80.42,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":53,"numerator":4825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"GDP.pdf","row_ids":["b8e5730fd93ac21867d62c37b5d60e40"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4350},"family":"AA-Omniscience","row_ids":["a92d6d21339d4ee97a6fa90820847da1","211e11b4d985ead9c997be1fc30d7961"],"score":82.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418}],"score":91.04,"weight":20}],"eligible":true,"id":"m-21b38c9ea818aebe","in_window":true,"name":"Muse Wonder Full","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":5,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0caa25aeefa10afc49082f57afb1971b","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"0.1","peer_count":53,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.371428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder-full:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.371428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e26762c5adc39cb8336491e6db427a8","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"Diamond","peer_count":64,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.485,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-wonder-full:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.485},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f2c7d3b6f687d90b18b5ecae64664442","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2 Hard","peer_count":52,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.565,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder-full:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.565},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fa97f8f99150ab638ce3d789e781973c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v2.1","peer_count":82,"percentile":97.53,"percentile_exact":{"denominator":81,"numerator":7900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1857.09,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder-full:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1857.09},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Wonder Full / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"07f1cdf7bbcb0da17ca1fb2ddf289fbe","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"4.0","peer_count":49,"percentile":88.54,"percentile_exact":{"denominator":24,"numerator":2125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":58.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"58.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row7","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":58.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d53e1ef4d2e9c319b7f0dacff98d83c","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7133074973223975,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder-full:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7133074973223975},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f4f6c3a22e75b41412d7850bc4b9cf5","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":78,"percentile":97.4,"percentile_exact":{"denominator":77,"numerator":7500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1729.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder-full:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1729.32},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"36981a7ad16ac52e29b5d161b5c3a617","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"2.0","peer_count":50,"percentile":87.76,"percentile_exact":{"denominator":49,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.377576262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder-full:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.377576262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder Full / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3d507b92bee70e08544b9eb65b20267a","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"1.1","peer_count":27,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":62.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"62.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row8"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row8","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":62.79},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b8e5730fd93ac21867d62c37b5d60e40","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"current AA protocol","peer_count":77,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.326,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder-full:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.326},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a92d6d21339d4ee97a6fa90820847da1","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.4096666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4096666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder Full (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"211e11b4d985ead9c997be1fc30d7961","included":true,"model_id":"m-21b38c9ea818aebe","original_version_label":"v1.1","peer_count":54,"percentile":94.34,"percentile_exact":{"denominator":53,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8243929983060418,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder-full:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder-full:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8243929983060418}],"release_date":"2026-10-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":88.51,"score_exact":{"denominator":5596519376448,"numerator":495344146904675},"source_id":"meta/muse-wonder-full"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":46512,"numerator":3838925},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":700},"family":"FrontierMath","row_ids":["a0152968ddf4e79954f370988a3aa877"],"score":77.78},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1775},"family":"Terminal-Bench-Science","row_ids":["0fd84fb90c507070cc93b2671c1dacd9"],"score":93.42},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":275},"family":"HLE","row_ids":["36082ffe886fbeb2647876856cf1e5c0"],"score":68.75},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4600},"family":"CritPt","row_ids":["57f29ac78c3749b38fa2eef65ef43c04"],"score":90.2}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","peer_count":52,"percentile":90.2,"percentile_exact":{"denominator":51,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59}],"score":82.54,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":35397,"numerator":3458243},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2600},"family":"GDPval-AA","row_ids":["c0ede537ba74a003e6de650e7903fd9e"],"score":96.3},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":292},"family":"Terminal-Bench","row_ids":["620eabc34583900c0a5610d2452e8c0e"],"score":97.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1800},"family":"AutomationBench","row_ids":["7e4dabafdeb0615d7fe8c2c4159042bd"],"score":94.74},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"AA-Briefcase","row_ids":["609c8649b3a04b44b0d6807fdfe15427"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":100},"family":"OSWorld","row_ids":["e0c59ec207d44193213f25c93cf18f5b"],"score":100.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2250},"family":"AA-AnalystAgent","row_ids":["b14e1b2737d16e7bb80a947abb8a320a"],"score":97.83}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","peer_count":82,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"4.0","peer_count":76,"percentile":97.33,"percentile_exact":{"denominator":3,"numerator":292},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":78,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"score":97.7,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":49,"numerator":4750},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4600},"family":"FrontierCode","row_ids":["2bdac0e90d63fa507e78af95b1c832a5"],"score":93.88},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"Vals Code Migration","row_ids":["165b56e6a0826f1ed0dc82b37b0885fe"],"score":100.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.0","peer_count":50,"percentile":93.88,"percentile_exact":{"denominator":49,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"1.1","peer_count":27,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88}],"score":96.94,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":3278075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1675},"family":"GDP.pdf","row_ids":["7992d55958a238d712427a450a500af9"],"score":88.16},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":650},"family":"SimpleQA Verified","row_ids":["b5dfd8ffb656e8526bed64b86b168f2a"],"score":38.24},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3450},"family":"AA-Omniscience","row_ids":["911853403e469f557e0c742924a5bacb","553708168125ea365f4f1a89df474d93"],"score":65.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923}],"score":63.83,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["34ec823cb4bb85ff7965822a48a3f4ce"],"score":100.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6}],"score":100.0,"weight":10}],"eligible":true,"id":"m-d123610543e5f5ed","in_window":true,"name":"Claude Sonnet 5.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":6,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a0152968ddf4e79954f370988a3aa877","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.8048780487804879,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a778da792f4073eb00b318e7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"a778da792f4073eb00b318e7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.8048780487804879},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0fd84fb90c507070cc93b2671c1dacd9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"0.1.0","peer_count":39,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":47.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"47.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row4"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row4","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":47.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-sonnet-5-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"36082ffe886fbeb2647876856cf1e5c0","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"Diamond","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.5,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df3527d6ed785b2579d75935"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"df3527d6ed785b2579d75935","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.5},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"57f29ac78c3749b38fa2eef65ef43c04","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2 Hard","peer_count":52,"percentile":90.2,"percentile_exact":{"denominator":51,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.59,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.59},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c0ede537ba74a003e6de650e7903fd9e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v2.1","peer_count":82,"percentile":96.3,"percentile_exact":{"denominator":27,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1837.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1837.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"620eabc34583900c0a5610d2452e8c0e","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"4.0","peer_count":76,"percentile":97.33,"percentile_exact":{"denominator":3,"numerator":292},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.636363636363636,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.636363636363636},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e4dabafdeb0615d7fe8c2c4159042bd","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7176882128193995,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7176882128193995},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"609c8649b3a04b44b0d6807fdfe15427","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":78,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1813.33,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1813.33},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"e0c59ec207d44193213f25c93cf18f5b","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":83.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"83.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":83.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b14e1b2737d16e7bb80a947abb8a320a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-sonnet-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2bdac0e90d63fa507e78af95b1c832a5","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"2.0","peer_count":50,"percentile":93.88,"percentile_exact":{"denominator":49,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.416993434343434,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.416993434343434},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"165b56e6a0826f1ed0dc82b37b0885fe","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"1.1","peer_count":27,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":86.88,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"86.88%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row1"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row1","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":86.88},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7992d55958a238d712427a450a500af9","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"current AA protocol","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.258,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.258},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"b5dfd8ffb656e8526bed64b86b168f2a","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"run 1.2.0","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.465,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-29","result_timestamp":"2026-09-29T15:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0e4f46c44c03ac96f0817efe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0e4f46c44c03ac96f0817efe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.465},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"911853403e469f557e0c742924a5bacb","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5146666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5146666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Sonnet 5.5 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"553708168125ea365f4f1a89df474d93","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5673076923076923,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5673076923076923},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Sonnet 5.5 For reference","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"34ec823cb4bb85ff7965822a48a3f4ce","included":true,"model_id":"m-d123610543e5f5ed","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":61.6,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"61.6% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column5","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":61.6}],"release_date":"2026-09-28","release_status":"published","release_verified":true,"result_updated_at":"2026-10-07","score":87.21,"score_exact":{"denominator":100015497792,"numerator":8722585306847},"source_id":"anthropic/claude-sonnet-5-5"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":15912,"numerator":1118525},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":500},"family":"FrontierMath","row_ids":["a21e66eb62f2bfca72e7c625965b008a"],"score":55.56},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1050},"family":"Terminal-Bench-Science","row_ids":["161dc1424fa043da938a084c4e822aa1"],"score":80.77},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":125},"family":"HLE","row_ids":["c73228fc930f557212149ca3c921d3fb"],"score":62.5},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1400},"family":"CritPt","row_ids":["024c0dcfccf03a6bb518b6c682477674"],"score":82.35}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"0.1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Diamond","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","peer_count":52,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505}],"score":70.29,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":21649031},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7700},"family":"GDPval-AA","row_ids":["5449b5fbd439528611d565a71ad28a45"],"score":95.06},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":262},"family":"Terminal-Bench","row_ids":["6fe9193a0e51ae328fa8f19ad476b5c1"],"score":87.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1700},"family":"AutomationBench","row_ids":["b2ddde3d390d97ad479920a7865821c3"],"score":89.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7200},"family":"AA-Briefcase","row_ids":["849c220b415c509e507a4751daf48b0a"],"score":93.51}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2.1","peer_count":82,"percentile":95.06,"percentile_exact":{"denominator":81,"numerator":7700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":78,"percentile":93.51,"percentile_exact":{"denominator":77,"numerator":7200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65}],"score":91.34,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":51000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3700},"family":"FrontierCode","row_ids":["3733f77f03c95d14f61b1419adf05cd6"],"score":75.51},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1100},"family":"Vals Code Migration","row_ids":["ab784b677fcf21bc1b248358effdf278"],"score":84.62}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"2.0","peer_count":50,"percentile":75.51,"percentile_exact":{"denominator":49,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"1.1","peer_count":27,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36}],"score":80.06,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":85700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1800},"family":"GDP.pdf","row_ids":["a4934e52e2688102b1f21d91ca098935"],"score":94.74},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4000},"family":"AA-Omniscience","row_ids":["3f07a1a0073c74e060ddc96a3f9e8bcb","58f09829d0f39ffd8664ffd449b9bfbd"],"score":75.47}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837}],"score":85.1,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":37775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2300},"family":"Humanity Sixth Sense","row_ids":["b9ee5aa138e25de06492fbd597ad313a"],"score":85.19},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1350},"family":"MMMU-Pro","row_ids":["bc79e5409c84a79a9897609e9d7ac22e"],"score":79.41}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","peer_count":28,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","peer_count":35,"percentile":79.41,"percentile_exact":{"denominator":17,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341}],"score":82.3,"weight":10}],"eligible":true,"id":"m-68480fc0423c5034","in_window":true,"name":"Muse Wonder","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":7,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-wonder-1.0_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a21e66eb62f2bfca72e7c625965b008a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7073170731707317,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T19:00:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c9cdf968af70c5e12940d5f7"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"c9cdf968af70c5e12940d5f7","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7073170731707317},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"161dc1424fa043da938a084c4e822aa1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"0.1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.257142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-wonder:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-wonder-1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"c73228fc930f557212149ca3c921d3fb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"Diamond","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":37.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f49d84bff87d1610e82482e4"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f49d84bff87d1610e82482e4","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":37.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"024c0dcfccf03a6bb518b6c682477674","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2 Hard","peer_count":52,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.505,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-wonder:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5449b5fbd439528611d565a71ad28a45","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v2.1","peer_count":82,"percentile":95.06,"percentile_exact":{"denominator":81,"numerator":7700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1770.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-wonder:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1770.34},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fe9193a0e51ae328fa8f19ad476b5c1","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-wonder:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b2ddde3d390d97ad479920a7865821c3","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6809455131207758,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-wonder:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6809455131207758},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"849c220b415c509e507a4751daf48b0a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":78,"percentile":93.51,"percentile_exact":{"denominator":77,"numerator":7200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1658.65,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-wonder:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1658.65},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3733f77f03c95d14f61b1419adf05cd6","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"2.0","peer_count":50,"percentile":75.51,"percentile_exact":{"denominator":49,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.27875303030303,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-wonder:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27875303030303},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Wonder / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"ab784b677fcf21bc1b248358effdf278","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"1.1","peer_count":27,"percentile":84.62,"percentile_exact":{"denominator":13,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":72.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"72.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row5","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":72.36},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a4934e52e2688102b1f21d91ca098935","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current AA protocol","peer_count":77,"percentile":94.74,"percentile_exact":{"denominator":19,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.306,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-wonder:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.306},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f07a1a0073c74e060ddc96a3f9e8bcb","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.38033333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.38033333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Wonder (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"58f09829d0f39ffd8664ffd449b9bfbd","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7912856374394837,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-wonder:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-wonder:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7912856374394837},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Wonder 1.0 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b9ee5aa138e25de06492fbd597ad313a","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current public protocol","peer_count":28,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":43.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b8e342aac48986ee66864044"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"b8e342aac48986ee66864044","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":43.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Wonder (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"bc79e5409c84a79a9897609e9d7ac22e","included":true,"model_id":"m-68480fc0423c5034","original_version_label":"current Standard 10","peer_count":35,"percentile":79.41,"percentile_exact":{"denominator":17,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838150289017341}],"release_date":"2026-10-12","release_status":"published","release_verified":true,"result_updated_at":"2026-10-12","score":81.67,"score_exact":{"denominator":310917743136,"numerator":25393529124989},"source_id":"meta/muse-wonder-1.0"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":41769,"numerator":3441925},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1075},"family":"Terminal-Bench-Science","row_ids":["440fdb6baf103b127060630dee151d5e"],"score":82.69},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5300},"family":"HLE","row_ids":["8cb18c97abf2e38058caf77fed29461c"],"score":84.13},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4100},"family":"CritPt","row_ids":["e4f3938369b9a55c0c1841556cc18184"],"score":80.39}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"0.1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","peer_count":64,"percentile":84.13,"percentile_exact":{"denominator":63,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","peer_count":52,"percentile":80.39,"percentile_exact":{"denominator":51,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495}],"score":82.4,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2725569,"numerator":238699840},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":8000},"family":"GDPval-AA","row_ids":["948bda542167bd107b882e3640f83a0c"],"score":98.77},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":100},"family":"Terminal-Bench","row_ids":["1c21aa90c01783fcd785c617ba4d90f9"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1850},"family":"AutomationBench","row_ids":["97fcee40bdfab50663d8d9aa18bd9131"],"score":97.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7400},"family":"AA-Briefcase","row_ids":["2b512abdb78eff5062c8cbfe38f0a197"],"score":96.1},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1050},"family":"AA-AnalystAgent","row_ids":["8b93841abeaaa5b29864bb557e6a9da7"],"score":45.65}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","peer_count":82,"percentile":98.77,"percentile_exact":{"denominator":81,"numerator":8000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"4.0","peer_count":76,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":78,"percentile":96.1,"percentile_exact":{"denominator":77,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"score":87.58,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":7,"numerator":500},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":500},"family":"FrontierCode","row_ids":["2cd7a142e0f06d143bb474c89fce8084"],"score":71.43}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"2.0","peer_count":50,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444}],"score":71.43,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":165875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1875},"family":"GDP.pdf","row_ids":["24d56f9893e1402bb483f6bef166ba08"],"score":98.68},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3500},"family":"AA-Omniscience","row_ids":["e4c0067c2ce56fc8c2df02c769685194","faf92c6808beb316ccdbea4a34863bce"],"score":66.04}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008}],"score":82.36,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1400},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1400},"family":"MMMU-Pro","row_ids":["75435cf97a5f0d98b110b2544004171f"],"score":82.35}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422}],"score":82.35,"weight":10}],"eligible":true,"id":"m-66c1aae520b7679a","in_window":true,"name":"Kimi K3.1 Pro","official_evidence":{"artifact_sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","basis":"official_release_page","date":"2026-10-24","sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","url":"https://www.kimi.com/en/blog/"},"primary_profile_id":"native","profile_id":"native","rank":8,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"440fdb6baf103b127060630dee151d5e","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"0.1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.342857142857143,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"kimi-k3-1-pro:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342857142857143},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8cb18c97abf2e38058caf77fed29461c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"Diamond","peer_count":64,"percentile":84.13,"percentile_exact":{"denominator":63,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.373,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k3-1-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e4f3938369b9a55c0c1841556cc18184","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2 Hard","peer_count":52,"percentile":80.39,"percentile_exact":{"denominator":51,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.495,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"kimi-k3-1-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.495},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"948bda542167bd107b882e3640f83a0c","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v2.1","peer_count":82,"percentile":98.77,"percentile_exact":{"denominator":81,"numerator":8000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1858.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k3-1-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1858.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c21aa90c01783fcd785c617ba4d90f9","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"4.0","peer_count":76,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.671717171717172,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k3-1-pro:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.671717171717172},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"97fcee40bdfab50663d8d9aa18bd9131","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":97.37,"percentile_exact":{"denominator":19,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7329771090921112,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k3-1-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7329771090921112},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b512abdb78eff5062c8cbfe38f0a197","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":78,"percentile":96.1,"percentile_exact":{"denominator":77,"numerator":7400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1674.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k3-1-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1674.17},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8b93841abeaaa5b29864bb557e6a9da7","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"kimi-k3-1-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2cd7a142e0f06d143bb474c89fce8084","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"2.0","peer_count":50,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.254519444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"kimi-k3-1-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.254519444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"24d56f9893e1402bb483f6bef166ba08","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current AA protocol","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.314,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k3-1-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e4c0067c2ce56fc8c2df02c769685194","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.48933333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48933333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Kimi K3.1 Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"faf92c6808beb316ccdbea4a34863bce","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6266318537859008,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"kimi-k3-1-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6266318537859008},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K3.1 Pro","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"75435cf97a5f0d98b110b2544004171f","included":true,"model_id":"m-66c1aae520b7679a","original_version_label":"current Standard 10","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.838728323699422,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k3-1-pro:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"kimi-k3-1-pro:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838728323699422}],"release_date":"2026-10-24","release_status":"published","release_verified":true,"result_updated_at":null,"score":81.49,"score_exact":{"denominator":127698358788,"numerator":10405967280655},"source_id":"moonshotai/kimi-k3.1-pro"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":3393175},"families":[{"benchmark_version":"2","exact":{"denominator":3,"numerator":200},"family":"FrontierMath","row_ids":["1004f40613fd7110bf20d6c47c5dcedb"],"score":66.67},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1700},"family":"Terminal-Bench-Science","row_ids":["50ddd51f0edebe68ad08c74703710c38"],"score":89.47},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5800},"family":"HLE","row_ids":["b185a4fb3502b3f94e9e93582e0d93a3"],"score":92.06},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1450},"family":"CritPt","row_ids":["3130f94996ac2afcc2f6ea8e78d8c87e"],"score":85.29}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","peer_count":39,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","peer_count":64,"percentile":92.06,"percentile_exact":{"denominator":63,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":83.37,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":16353414,"numerator":1123315619},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6500},"family":"GDPval-AA","row_ids":["feff6b1ffb0f6cb8ff2773defec53d29"],"score":80.25},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":278},"family":"Terminal-Bench","row_ids":["58499c853c2644c7eb9a4bc16fc208c6"],"score":92.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1875},"family":"AutomationBench","row_ids":["a2113bd24aca42a48eba4972dbb6e6d4"],"score":98.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5800},"family":"AA-Briefcase","row_ids":["782453050df6541e218363a18f9bb044"],"score":75.32},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"OSWorld","row_ids":["f61750e8ba74bd2c7928dd7c56124cb4"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1500},"family":"AA-AnalystAgent","row_ids":["4f1262f25c1f33cd755aebe2e08018dd"],"score":65.22}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","peer_count":82,"percentile":80.25,"percentile_exact":{"denominator":81,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","peer_count":76,"percentile":92.67,"percentile_exact":{"denominator":3,"numerator":278},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":78,"percentile":75.32,"percentile_exact":{"denominator":77,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","peer_count":24,"percentile":65.22,"percentile_exact":{"denominator":23,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875}],"score":68.69,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":1911,"numerator":161950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3600},"family":"FrontierCode","row_ids":["aabef6305c7bf9e588cc8a4221b8394d"],"score":73.47},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1050},"family":"Vals Code Migration","row_ids":["2905dc672ed8db458e1dc8c5a09c04a7"],"score":80.77},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":100},"family":"MirrorCode","row_ids":["d2afdbf96101f2fb4499bd4ee1fc2881"],"score":100.0}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","peer_count":50,"percentile":73.47,"percentile_exact":{"denominator":49,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","peer_count":27,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889}],"score":84.75,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":2014,"numerator":175925},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1575},"family":"GDP.pdf","row_ids":["da4cc7b10c537dba69aca5448e60f8f7"],"score":82.89},{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":100},"family":"SimpleQA Verified","row_ids":["56d131d2a5cbdffe166bdda67e598654"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"MultiChallenge","row_ids":["2db7888810c9c42969f9152c713f9d6e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4850},"family":"AA-Omniscience","row_ids":["0588c4d7908c2d87deb09a5a03cbe1c3","2f9d0f30b6d18925324be8097217f20f"],"score":91.51}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383}],"score":87.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":2,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["18175febe04f7246f991614fc6ca93f9"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"Chartography","row_ids":["13ad1c51a1b24d02661158660b44e5e0"],"score":75.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","peer_count":28,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6}],"score":87.5,"weight":10}],"eligible":true,"id":"m-7ff6a9f995dcaa1c","in_window":true,"name":"Gemini 4 Argon","official_evidence":{"artifact_sha256":"555ea742fc3bcfcd343ab416bbcdf209448d15c9f06f0240eb19c9875fdccf00","basis":"official_release_page","date":"2026-10-13","sha256":"555ea742fc3bcfcd343ab416bbcdf209448d15c9f06f0240eb19c9875fdccf00","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-argon/"},"primary_profile_id":"native","profile_id":"native","rank":9,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1004f40613fd7110bf20d6c47c5dcedb","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.7804878048780488,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:14:57.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1223ccbd15d66ddf0ce74c9e"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"1223ccbd15d66ddf0ce74c9e","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.7804878048780488},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 4 Argon / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"50ddd51f0edebe68ad08c74703710c38","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"0.1.0","peer_count":39,"percentile":89.47,"percentile_exact":{"denominator":19,"numerator":1700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"44.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row5"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row5","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":44.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b185a4fb3502b3f94e9e93582e0d93a3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"Diamond","peer_count":64,"percentile":92.06,"percentile_exact":{"denominator":63,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.567,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-4-argon:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.567},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3130f94996ac2afcc2f6ea8e78d8c87e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-4-argon:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"feff6b1ffb0f6cb8ff2773defec53d29","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v2.1","peer_count":82,"percentile":80.25,"percentile_exact":{"denominator":81,"numerator":6500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.66,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-4-argon:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.66},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"58499c853c2644c7eb9a4bc16fc208c6","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"4.0","peer_count":76,"percentile":92.67,"percentile_exact":{"denominator":3,"numerator":278},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.570707070707071,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-4-argon:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.570707070707071},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a2113bd24aca42a48eba4972dbb6e6d4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":98.68,"percentile_exact":{"denominator":19,"numerator":1875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7751346945899195,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-4-argon:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7751346945899195},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"782453050df6541e218363a18f9bb044","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":78,"percentile":75.32,"percentile_exact":{"denominator":77,"numerator":5800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1485.03,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-4-argon:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1485.03},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld 2.1 offline subset","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"agents","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"f61750e8ba74bd2c7928dd7c56124cb4","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official comparison image; OSWorld 2.1 offline subset; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"ab44942b390af73e564b6930","raw_value":74.7,"reference_revision":"1ff527363d985d3127580f612aad5b4b5714db22db55079be25cab83a9248830","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"74.7%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:OSWorld:OSWorld 2.1 offline subset","source_record_id":"carbon-image:OSWorld 2.1 offline subset:Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":74.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4f1262f25c1f33cd755aebe2e08018dd","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA pass^5","peer_count":24,"percentile":65.22,"percentile_exact":{"denominator":23,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-4-argon:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"aabef6305c7bf9e588cc8a4221b8394d","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"2.0","peer_count":50,"percentile":73.47,"percentile_exact":{"denominator":49,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.257076262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-4-argon:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257076262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 4 Argon / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2905dc672ed8db458e1dc8c5a09c04a7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"1.1","peer_count":27,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":70.8,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"70.80%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row6","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":70.8},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d2afdbf96101f2fb4499bd4ee1fc2881","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 0.1.200","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.5888888888888889,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-14","result_timestamp":"2026-10-14T10:00:34.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57b82d61a96d2570862b310c"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"57b82d61a96d2570862b310c","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5888888888888889},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"da4cc7b10c537dba69aca5448e60f8f7","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current AA protocol","peer_count":77,"percentile":82.89,"percentile_exact":{"denominator":19,"numerator":1575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.218,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-4-argon:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.218},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-4-argon_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"56d131d2a5cbdffe166bdda67e598654","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"run 1.2.0","peer_count":35,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.76,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T17:15:14.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b86904d16e148c514d0e38ab"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"b86904d16e148c514d0e38ab","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.76},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Gemini 4 Argon","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"2db7888810c9c42969f9152c713f9d6e","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current LLM protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":78.7,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57617fce725f614ee6ffcdb6"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"57617fce725f614ee6ffcdb6","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":78.7},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0588c4d7908c2d87deb09a5a03cbe1c3","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.483,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.483},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 4 Argon (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2f9d0f30b6d18925324be8097217f20f","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.9574468085106383,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-4-argon:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-4-argon:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9574468085106383},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-4-Argon (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"18175febe04f7246f991614fc6ca93f9","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current public protocol","peer_count":28,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":55.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8041bb45d96cc4070f198ff1"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8041bb45d96cc4070f198ff1","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":55.4},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography ","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:15.865489+00:00","component_key":"score","configuration":"Gemini 4 Argon / official Gemini 4 Carbon comparison","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-image-be7a254ccb1504f5","id":"13ad1c51a1b24d02661158660b44e5e0","included":true,"model_id":"m-7ff6a9f995dcaa1c","original_version_label":"current published dataset","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official comparison image; Chartography ; original image and release footnotes preserved. Offline-subset and AA-labelled protocols remain distinct.","protocol_id":"bc37dbdbfe633db52e1c0b8e","raw_value":71.6,"reference_revision":"3a4e8c2cc1eaa2dd3ffa0f26bedff2a2146ad75b68527a621105092727179481","research_artifact_sha256":"47d5f256ee73b6eaf7412ffd81ad9679727448ef6c6bb2d6dd6a32ade6c452b6","response_sha256":"efefd16761a463db5a23769b8c2d85ea2873d4adc9e7f2563728a620d5501079","result_date":"2026-10-28","source_cell":"71.6%","source_id":"s-ace8373a102d94a857d5","source_kind":"official_release_comparison","source_locator":{"image_url":"https://storage.googleapis.com/gweb-uniblog-publish-prod/images/gemini_4_carbon__evals-table.width-2000.format-webp.webp","record":"carbon-image:Chartography :Gemini 4 Argon"},"source_name":"blog.google","source_protocol_key":"official:blog.google:Chartography:Chartography","source_record_id":"carbon-image:Chartography :Gemini 4 Argon","source_url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/","unit":"%","value":71.6}],"release_date":"2026-10-13","release_status":"published","release_verified":true,"result_updated_at":"2026-10-28","score":81.19,"score_exact":{"denominator":2681665534548,"numerator":217712535238087},"source_id":"google/gemini-4-argon"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":31824,"numerator":2630125},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":800},"family":"FrontierMath","row_ids":["428bcb6f2312f54ea131660ec5510f0c"],"score":88.89},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":1200},"family":"Terminal-Bench-Science","row_ids":["3a64ac33a8e0bec7c9bbb47064e6200b"],"score":92.31},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":225},"family":"HLE","row_ids":["9c15f08ff3f1d95aac1ab4460a6a090b"],"score":56.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4750},"family":"CritPt","row_ids":["624fa427f469bd9c8115ca8491b3f8a8"],"score":93.14}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","peer_count":53,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68}],"score":82.65,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":17835809},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5600},"family":"GDPval-AA","row_ids":["b0fdd931f8b81700b478f4129dec593b"],"score":69.14},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":268},"family":"Terminal-Bench","row_ids":["35604c8a1315056deeba470a2491f933"],"score":89.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1450},"family":"AutomationBench","row_ids":["1b6722ff89e791988385d9c44034ffe3"],"score":76.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5100},"family":"AA-Briefcase","row_ids":["07c0ef81ba075e768f25e1177f4ce283"],"score":66.23}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","peer_count":82,"percentile":69.14,"percentile_exact":{"denominator":81,"numerator":5600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","peer_count":76,"percentile":89.33,"percentile_exact":{"denominator":3,"numerator":268},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":78,"percentile":66.23,"percentile_exact":{"denominator":77,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36}],"score":75.25,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":4700},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4700},"family":"FrontierCode","row_ids":["bf241286cb096d94e2aa92889567c092"],"score":95.92}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","peer_count":50,"percentile":95.92,"percentile_exact":{"denominator":49,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505}],"score":95.92,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":9124175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1775},"family":"GDP.pdf","row_ids":["5a11c32835ae26c4dc93dd63135b1a2a"],"score":93.42},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1150},"family":"SimpleQA Verified","row_ids":["52a58ef3bf54eb42e218e3c094aec598"],"score":67.65},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"MultiChallenge","row_ids":["c65e268d91030372566c3ec5c7a1ab64"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3600},"family":"AA-Omniscience","row_ids":["43cafc16779c4f9a017fe8ae9ccdcb81","5ac877348d0fb5ad33d4ed61b4154929"],"score":67.92}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129}],"score":66.62,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1450},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1450},"family":"MMMU-Pro","row_ids":["cd4afdb9196fc47218c2c1085364d08a"],"score":85.29}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555}],"score":85.29,"weight":10}],"eligible":true,"id":"m-1dcd13c2fce7d97b","in_window":true,"name":"GPT-6.1 Terra","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":10,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"428bcb6f2312f54ea131660ec5510f0c","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.926829268292683,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b7360141e1e95a4ad46f47cb"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"b7360141e1e95a4ad46f47cb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.926829268292683},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3a64ac33a8e0bec7c9bbb47064e6200b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"0.1","peer_count":53,"percentile":92.31,"percentile_exact":{"denominator":13,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.576190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-1-terra:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.576190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6.1-terra","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"9c15f08ff3f1d95aac1ab4460a6a090b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"Diamond","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":35.7,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f33a41071a557e4fe07d7540"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f33a41071a557e4fe07d7540","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":35.7},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"624fa427f469bd9c8115ca8491b3f8a8","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2 Hard","peer_count":52,"percentile":93.14,"percentile_exact":{"denominator":51,"numerator":4750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.68,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-1-terra:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.68},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0fdd931f8b81700b478f4129dec593b","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v2.1","peer_count":82,"percentile":69.14,"percentile_exact":{"denominator":81,"numerator":5600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1475.63,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-1-terra:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.63},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"35604c8a1315056deeba470a2491f933","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"4.0","peer_count":76,"percentile":89.33,"percentile_exact":{"denominator":3,"numerator":268},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-6-1-terra:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b6722ff89e791988385d9c44034ffe3","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6016762327257726,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-1-terra:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6016762327257726},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07c0ef81ba075e768f25e1177f4ce283","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":78,"percentile":66.23,"percentile_exact":{"denominator":77,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1401.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-1-terra:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.36},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"bf241286cb096d94e2aa92889567c092","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"2.0","peer_count":50,"percentile":95.92,"percentile_exact":{"denominator":49,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.423200505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-1-terra:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423200505050505},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5a11c32835ae26c4dc93dd63135b1a2a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current AA protocol","peer_count":77,"percentile":93.42,"percentile_exact":{"denominator":19,"numerator":1775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.286,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-1-terra:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6.1-terra_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"52a58ef3bf54eb42e218e3c094aec598","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"run 1.2.0","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.573,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-26","result_timestamp":"2026-10-26T20:00:37.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2dff251325a5cd80a1750827"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2dff251325a5cd80a1750827","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.573},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"GPT 6.1 Terra","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"c65e268d91030372566c3ec5c7a1ab64","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1f39f348d076cd378782e59f"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"1f39f348d076cd378782e59f","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"43cafc16779c4f9a017fe8ae9ccdcb81","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47633333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47633333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6.1 Terra (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac877348d0fb5ad33d4ed61b4154929","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6549968173138129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-1-terra:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6549968173138129},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6.1 Terra (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"cd4afdb9196fc47218c2c1085364d08a","included":true,"model_id":"m-1dcd13c2fce7d97b","original_version_label":"current Standard 10","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.846820809248555,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-1-terra:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-1-terra:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.846820809248555}],"release_date":"2026-10-26","release_status":"published","release_verified":true,"result_updated_at":"2026-10-26","score":80.51,"score_exact":{"denominator":621835486272,"numerator":50065762899577},"source_id":"openai/gpt-6.1-terra"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":533596,"numerator":44974135},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":500},"family":"FrontierMath","row_ids":["a450c91f279ff648cb2b208a74552ba0"],"score":71.43},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5700},"family":"GPQA Diamond","row_ids":["0539c349124105a728b2ffaeadc2dabd"],"score":96.61},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1650},"family":"Terminal-Bench-Science","row_ids":["2012b47ec868b52e455d3d8d66a1a48a"],"score":86.84},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":325},"family":"HLE","row_ids":["16b64e0ecfeb64ebd3bc40ba58f5f691"],"score":81.25},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1450},"family":"CritPt","row_ids":["2a25d966a3bebfa161181074ba7fbad0"],"score":85.29}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":96.61,"percentile_exact":{"denominator":59,"numerator":5700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","peer_count":39,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":84.28,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":496053558,"numerator":42907839551},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":2375},"family":"tau3-Banking","row_ids":["a5ec26f61468360fa82a5667cce554db"],"score":91.35},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7600},"family":"GDPval-AA","row_ids":["1bc96450387b9110981d343f4dfd20eb"],"score":93.83},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":262},"family":"Terminal-Bench","row_ids":["71f828c6aa1f518ef08599ca0a1b0570"],"score":87.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1400},"family":"AutomationBench","row_ids":["9a8fd23ce0ae510627402be3083ec808"],"score":73.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7300},"family":"AA-Briefcase","row_ids":["cab25ff85e887187f7f900705800483c"],"score":94.81},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"OSWorld","row_ids":["520b3d1c3a49b9ec75dd20571c69ae96"],"score":66.67},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2250},"family":"AA-AnalystAgent","row_ids":["172f756dc0c16d9248bcf51116651a4b"],"score":97.83}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","peer_count":82,"percentile":93.83,"percentile_exact":{"denominator":81,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":78,"percentile":94.81,"percentile_exact":{"denominator":77,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.1","peer_count":7,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575}],"score":86.5,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":98,"numerator":5325},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4100},"family":"FrontierCode","row_ids":["468e0160c503684573c12f005950e313"],"score":83.67},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["a9df38e2550ef9f8d566441fba3e6c66"],"score":25.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.0","peer_count":50,"percentile":83.67,"percentile_exact":{"denominator":49,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"1.1","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4}],"score":54.34,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":102714,"numerator":8141825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3425},"family":"GDP.pdf","row_ids":["5d9e8383609a67f5871ab595963f24e2"],"score":90.13},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1500},"family":"SimpleQA Verified","row_ids":["add598a99258de2861924fc25c825623"],"score":88.24},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3150},"family":"AA-Omniscience","row_ids":["b829940da89808936844666d294a78aa","2cbd1e01e64dad121dca7aaf300f12be"],"score":59.43}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996}],"score":79.27,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":9,"numerator":800},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":700},"family":"Humanity Sixth Sense","row_ids":["2d80b1607b40fc4d214cb22e259d1221"],"score":77.78},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"Chartography","row_ids":["a7fcafd577ada1f13084bdc5d25fc8ab"],"score":100.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","peer_count":28,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4}],"score":88.89,"weight":10}],"eligible":true,"id":"m-98ee7005cf45b532","in_window":true,"name":"Claude Fable 5.1","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":11,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"reasoning","direction":"higher_better","id":"a450c91f279ff648cb2b208a74552ba0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":87.8,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":87.8},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0539c349124105a728b2ffaeadc2dabd","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":96.61,"percentile_exact":{"denominator":59,"numerator":5700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.937373737373737,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"claude-fable-5-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.937373737373737},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5.1 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"2012b47ec868b52e455d3d8d66a1a48a","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"0.1.0","peer_count":39,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":40.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"40.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row6"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row6","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":40.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-fable-5-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"16b64e0ecfeb64ebd3bc40ba58f5f691","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"Diamond","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":51.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a6f3483331435f0013d4fb40"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"a6f3483331435f0013d4fb40","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":51.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2a25d966a3bebfa161181074ba7fbad0","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2 Hard","peer_count":52,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.525,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-fable-5-1:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5ec26f61468360fa82a5667cce554db","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1bc96450387b9110981d343f4dfd20eb","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v2.1","peer_count":82,"percentile":93.83,"percentile_exact":{"denominator":81,"numerator":7600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1763.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-fable-5-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1763.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"71f828c6aa1f518ef08599ca0a1b0570","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"4.0","peer_count":76,"percentile":87.33,"percentile_exact":{"denominator":3,"numerator":262},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.52020202020202,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-fable-5-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.52020202020202},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a8fd23ce0ae510627402be3083ec808","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5937591715646424,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-fable-5-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5937591715646424},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab25ff85e887187f7f900705800483c","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":78,"percentile":94.81,"percentile_exact":{"denominator":77,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1669.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-fable-5-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1669.49},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"520b3d1c3a49b9ec75dd20571c69ae96","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.1","peer_count":7,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":80.7,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"80.7% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":80.7},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"172f756dc0c16d9248bcf51116651a4b","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA pass^5","peer_count":24,"percentile":97.83,"percentile_exact":{"denominator":23,"numerator":2250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.575,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-fable-5-1:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"468e0160c503684573c12f005950e313","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"2.0","peer_count":50,"percentile":83.67,"percentile_exact":{"denominator":49,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.316248737373737,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-fable-5-1:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.316248737373737},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5.1","dimension":"coding","direction":"higher_better","id":"a9df38e2550ef9f8d566441fba3e6c66","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"1.1","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.4,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column3","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":67.4},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5d9e8383609a67f5871ab595963f24e2","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current AA protocol","peer_count":77,"percentile":90.13,"percentile_exact":{"denominator":38,"numerator":3425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.262,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-fable-5-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.262},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-fable-5-1_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"add598a99258de2861924fc25c825623","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"run 1.2.0","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.708,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-01","result_timestamp":"2026-09-01T18:31:59.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"530b4e278e6394bb8c0803eb"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"530b4e278e6394bb8c0803eb","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.708},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b829940da89808936844666d294a78aa","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.684,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.684},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Fable 5.1 (Max, Default Fallback)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2cbd1e01e64dad121dca7aaf300f12be","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29957805907172996,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-fable-5-1:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29957805907172996},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d80b1607b40fc4d214cb22e259d1221","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"current public protocol","peer_count":28,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":40.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"326cfb1d36289e8186786c0a"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"326cfb1d36289e8186786c0a","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":40.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Fable 5.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"a7fcafd577ada1f13084bdc5d25fc8ab","included":true,"model_id":"m-98ee7005cf45b532","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":88.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"88.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column2","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":88.4}],"release_date":"2026-09-01","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":78.31,"score_exact":{"denominator":30136812673968,"numerator":2359877379520649},"source_id":"anthropic/claude-fable-5-1"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3717,"numerator":264100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5200},"family":"GPQA Diamond","row_ids":["c6f5e59b8c457f8fd73cf7fe23e60786"],"score":88.14},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3400},"family":"HLE","row_ids":["1e856649fb09c4da785f8c4ea8ad05e0"],"score":53.97}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":88.14,"percentile_exact":{"denominator":59,"numerator":5200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond","peer_count":64,"percentile":53.97,"percentile_exact":{"denominator":63,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163}],"score":71.05,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":603860552},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1150},"family":"tau3-Banking","row_ids":["24d253e4fe3d6de3cc614f0a892166b1"],"score":88.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6700},"family":"GDPval-AA","row_ids":["c47c79a0ac9d03dc0858a3e32eb10cf4"],"score":82.72},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":68},"family":"Terminal-Bench","row_ids":["0978fbc73a68e2ee70d02c130e0d1cff"],"score":68.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1250},"family":"AutomationBench","row_ids":["12fa4c01c24c67c3a86dd523987aed83"],"score":65.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6700},"family":"AA-Briefcase","row_ids":["62ba5fa448ec9e7825ee5093f17e487d"],"score":87.01}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v2.1","peer_count":82,"percentile":82.72,"percentile_exact":{"denominator":81,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"4.0","peer_count":76,"percentile":68.0,"percentile_exact":{"denominator":1,"numerator":68},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":78,"percentile":87.01,"percentile_exact":{"denominator":77,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51}],"score":78.4,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["80a064d01bdbcc1f636799e9dd23518a"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"1.1","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7}],"score":100.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":6042,"numerator":401725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["1bcc725977832eb3960634353511e9be"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2525},"family":"GDP.pdf","row_ids":["046ca38af4c5ad3c8241882797259027"],"score":66.45},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1750},"family":"AA-Omniscience","row_ids":["0351ad1258b35881ec0b312bb72c8d29","8283a5fc47f30f74b98c14b10fde0469"],"score":33.02}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129}],"score":66.49,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1050},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1050},"family":"MMMU-Pro","row_ids":["42c55648590ad4a69cdc2a21de6faac7"],"score":61.76}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676}],"score":61.76,"weight":10}],"eligible":true,"id":"m-e1cba225d1430824","in_window":true,"name":"Qwen3.8-Flash-Next","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":12,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6f5e59b8c457f8fd73cf7fe23e60786","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":88.14,"percentile_exact":{"denominator":59,"numerator":5200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.923232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-flash-next:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.923232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e856649fb09c4da785f8c4ea8ad05e0","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"Diamond","peer_count":64,"percentile":53.97,"percentile_exact":{"denominator":63,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.163,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-flash-next:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.163},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d253e4fe3d6de3cc614f0a892166b1","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":88.46,"percentile_exact":{"denominator":13,"numerator":1150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.45360824742268,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-flash-next:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45360824742268},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c47c79a0ac9d03dc0858a3e32eb10cf4","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v2.1","peer_count":82,"percentile":82.72,"percentile_exact":{"denominator":81,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1635.78,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-flash-next:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1635.78},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0978fbc73a68e2ee70d02c130e0d1cff","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"4.0","peer_count":76,"percentile":68.0,"percentile_exact":{"denominator":1,"numerator":68},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.252525252525253,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-flash-next:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252525252525253},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"12fa4c01c24c67c3a86dd523987aed83","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.559083218951688,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-flash-next:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.559083218951688},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62ba5fa448ec9e7825ee5093f17e487d","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":78,"percentile":87.01,"percentile_exact":{"denominator":77,"numerator":6700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1573.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-flash-next:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1573.51},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"80a064d01bdbcc1f636799e9dd23518a","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"1.1","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":58.7,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"58.7","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":58.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"1bcc725977832eb3960634353511e9be","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current published protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":81.3,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"81.3","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":81.3},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"046ca38af4c5ad3c8241882797259027","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-flash-next:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0351ad1258b35881ec0b312bb72c8d29","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8-Flash-Next","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8283a5fc47f30f74b98c14b10fde0469","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.532258064516129,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-flash-next:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.532258064516129},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8-Flash-Next","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"42c55648590ad4a69cdc2a21de6faac7","included":true,"model_id":"m-e1cba225d1430824","original_version_label":"current Standard 10","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.797687861271676,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-flash-next:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-flash-next:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.797687861271676}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"result_updated_at":"2026-08-26","score":76.84,"score_exact":{"denominator":818935127010,"numerator":62923865665709},"source_id":"alibaba/qwen3.8-flash-next"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":27846,"numerator":2129875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1975},"family":"Terminal-Bench-Science","row_ids":["bb1f3609ad48f7662af1e4f3a15cef7a"],"score":75.96},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1700},"family":"HLE","row_ids":["5198f9fefccca8b5961dab3ccd43cfb2"],"score":80.95},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3700},"family":"CritPt","row_ids":["6c2b0cbfd16faf311bd4edca68e40d0e"],"score":72.55}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","peer_count":64,"percentile":80.95,"percentile_exact":{"denominator":21,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","peer_count":52,"percentile":72.55,"percentile_exact":{"denominator":51,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39}],"score":76.49,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1197,"numerator":95609},"families":[{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":248},"family":"Terminal-Bench","row_ids":["b9ebef353db3bd587384633039b7a5bb"],"score":82.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1625},"family":"AutomationBench","row_ids":["01aaf3c58f06d7a8962fe3fd60fa4286"],"score":85.53},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":500},"family":"AA-Briefcase","row_ids":["732f112b308243fb337093d80efff0c3"],"score":71.43}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","peer_count":76,"percentile":82.67,"percentile_exact":{"denominator":3,"numerator":248},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":85.53,"percentile_exact":{"denominator":19,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":78,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69}],"score":79.87,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":4000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":4000},"family":"FrontierCode","row_ids":["8ed6fd9401411cecf6b0e548f78e12a5"],"score":81.63}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","peer_count":50,"percentile":81.63,"percentile_exact":{"denominator":49,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101}],"score":81.63,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1007,"numerator":64350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1550},"family":"GDP.pdf","row_ids":["b84d512b0d7ca681d5568eba0c2a9675"],"score":81.58},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2450},"family":"AA-Omniscience","row_ids":["795265b8e61a4f96a23b56d372ad2050","7d74696c1e12470b21a043973afd962c"],"score":46.23}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356}],"score":63.9,"weight":20}],"eligible":true,"id":"m-b88efd1e16a0b031","in_window":true,"name":"Tencent Hy4","official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-11-03","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":13,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bb1f3609ad48f7662af1e4f3a15cef7a","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5198f9fefccca8b5961dab3ccd43cfb2","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"Diamond","peer_count":64,"percentile":80.95,"percentile_exact":{"denominator":21,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.327,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.327},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6c2b0cbfd16faf311bd4edca68e40d0e","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v2 Hard","peer_count":52,"percentile":72.55,"percentile_exact":{"denominator":51,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.39,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9ebef353db3bd587384633039b7a5bb","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"4.0","peer_count":76,"percentile":82.67,"percentile_exact":{"denominator":3,"numerator":248},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.373737373737374,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.373737373737374},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01aaf3c58f06d7a8962fe3fd60fa4286","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":85.53,"percentile_exact":{"denominator":19,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6561018180201071,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6561018180201071},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732f112b308243fb337093d80efff0c3","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":78,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1446.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1446.69},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"8ed6fd9401411cecf6b0e548f78e12a5","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"2.0","peer_count":50,"percentile":81.63,"percentile_exact":{"denominator":49,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.31447601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.31447601010101},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b84d512b0d7ca681d5568eba0c2a9675","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"current AA protocol","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.216,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"795265b8e61a4f96a23b56d372ad2050","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.49966666666666665,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.49966666666666665},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Tencent Hy4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d74696c1e12470b21a043973afd962c","included":true,"model_id":"m-b88efd1e16a0b031","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.10926049300466356,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10926049300466356}],"release_date":"2026-11-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":75.77,"score_exact":{"denominator":392572908,"numerator":29747186485},"source_id":"published/aa:hunyuan-dev1030-aa"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":294400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5600},"family":"GPQA Diamond","row_ids":["b2e2268e4647b383f382e5a7860dafa3"],"score":94.92},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4000},"family":"HLE","row_ids":["b6295a8854af11f8a1045539e4932330"],"score":63.49}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":94.92,"percentile_exact":{"denominator":59,"numerator":5600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond","peer_count":64,"percentile":63.49,"percentile_exact":{"denominator":63,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201}],"score":79.2,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":10783773,"numerator":844623034},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1275},"family":"tau3-Banking","row_ids":["960dd18162897dbffed3123d8b9728ee"],"score":98.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6200},"family":"GDPval-AA","row_ids":["fcdd405fcadd197dbec16caec46c9124"],"score":76.54},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":56},"family":"Terminal-Bench","row_ids":["8981298b68f7a13ad078d01ac54f2b30"],"score":56.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1325},"family":"AutomationBench","row_ids":["f2b687ab9797dca77a0de360cefeb5cf"],"score":69.74},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5400},"family":"AA-Briefcase","row_ids":["be21483d79b65c8a6ad01a6eee1e0103"],"score":70.13},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"EnterpriseOps-Gym","row_ids":["32758e4ec324e0ec55fbc6f952cf8937"],"score":88.89},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":800},"family":"MCP Atlas","row_ids":["7b2615183b2725bb4693774f30af4a5b"],"score":88.89}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":98.08,"percentile_exact":{"denominator":13,"numerator":1275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v2.1","peer_count":82,"percentile":76.54,"percentile_exact":{"denominator":81,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"4.0","peer_count":76,"percentile":56.0,"percentile_exact":{"denominator":1,"numerator":56},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":78,"percentile":70.13,"percentile_exact":{"denominator":77,"numerator":5400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA run","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5}],"score":78.32,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":238325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2525},"family":"GDP.pdf","row_ids":["93e5cef565694d0b6ece46832b2fc2e8"],"score":66.45},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2750},"family":"AA-Omniscience","row_ids":["872cbb8006efe075a4c9b62b4f118011","7f66b653a79b36bb8c9fbcd69744e5f5"],"score":51.89}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656}],"score":59.17,"weight":20}],"eligible":true,"id":"m-fb920161af1862b7","in_window":true,"name":"Qwen3.8 2.4T A95B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":14,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b2e2268e4647b383f382e5a7860dafa3","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":94.92,"percentile_exact":{"denominator":59,"numerator":5600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.935353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-2-4t-a95b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.935353535353535},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b6295a8854af11f8a1045539e4932330","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"Diamond","peer_count":64,"percentile":63.49,"percentile_exact":{"denominator":63,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.201,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.201},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"960dd18162897dbffed3123d8b9728ee","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":98.08,"percentile_exact":{"denominator":13,"numerator":1275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.490721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-2-4t-a95b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.490721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fcdd405fcadd197dbec16caec46c9124","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v2.1","peer_count":82,"percentile":76.54,"percentile_exact":{"denominator":81,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1615.72,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-2-4t-a95b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1615.72},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8981298b68f7a13ad078d01ac54f2b30","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"4.0","peer_count":76,"percentile":56.0,"percentile_exact":{"denominator":1,"numerator":56},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.111111111111111,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-2-4t-a95b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.111111111111111},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2b687ab9797dca77a0de360cefeb5cf","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.572473765458222,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-2-4t-a95b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.572473765458222},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"be21483d79b65c8a6ad01a6eee1e0103","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":78,"percentile":70.13,"percentile_exact":{"denominator":77,"numerator":5400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1430.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-2-4t-a95b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1430.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32758e4ec324e0ec55fbc6f952cf8937","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA run","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4735899731423456,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-2-4t-a95b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4735899731423456},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Qwen3.8-2.4T-A95B (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"7b2615183b2725bb4693774f30af4a5b","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.5,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"fa04d48e235ad15f143a5cbe"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"fa04d48e235ad15f143a5cbe","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"93e5cef565694d0b6ece46832b2fc2e8","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"current AA protocol","peer_count":77,"percentile":66.45,"percentile_exact":{"denominator":38,"numerator":2525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.156,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-2-4t-a95b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.156},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"872cbb8006efe075a4c9b62b4f118011","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":47.17,"percentile_exact":{"denominator":53,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.314,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.314},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 2.4T A95B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f66b653a79b36bb8c9fbcd69744e5f5","included":true,"model_id":"m-fb920161af1862b7","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6554907677356656,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-2-4t-a95b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-2-4t-a95b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6554907677356656}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":73.16,"score_exact":{"denominator":472092014394,"numerator":34540416783815},"source_id":"alibaba/qwen3.8-2.4t-a95b"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":5967,"numerator":413600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":800},"family":"Terminal-Bench-Science","row_ids":["7ead6883bce4c045cf26f5b116b4cd48"],"score":61.54},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":700},"family":"HLE","row_ids":["89d939473fcae282b052740fc3fcfd7a"],"score":77.78},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3500},"family":"CritPt","row_ids":["34dffe7c2a7d437da340dfb5246fe9c8"],"score":68.63}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","peer_count":64,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","peer_count":52,"percentile":68.63,"percentile_exact":{"denominator":51,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"score":69.31,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":35432397,"numerator":2557154263},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1050},"family":"tau3-Banking","row_ids":["66284fe19cdf80bb97a995a814e23753"],"score":80.77},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6400},"family":"GDPval-AA","row_ids":["f53e9f08adf2987f0e2c59d35bf57232"],"score":79.01},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":184},"family":"Terminal-Bench","row_ids":["b334bad67fbd4a96eed8e669e0111fbb"],"score":61.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1025},"family":"AutomationBench","row_ids":["fb9e5b876d81fca5e8cf53e8169709cb"],"score":53.95},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":900},"family":"AA-Briefcase","row_ids":["ae0bf141fcba5d508b0da173dcf01538"],"score":81.82},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":650},"family":"EnterpriseOps-Gym","row_ids":["ccd3047c248ced9697a972d3c394d08d"],"score":72.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1750},"family":"AA-AnalystAgent","row_ids":["e23e94bce569e3668ea49a1736f85e07"],"score":76.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","peer_count":82,"percentile":79.01,"percentile_exact":{"denominator":81,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","peer_count":76,"percentile":61.33,"percentile_exact":{"denominator":3,"numerator":184},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":78,"percentile":81.82,"percentile_exact":{"denominator":11,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125}],"score":72.17,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":3400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3400},"family":"FrontierCode","row_ids":["0fe380f78c53ebf4d268a83df481e0ad"],"score":69.39}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","peer_count":50,"percentile":69.39,"percentile_exact":{"denominator":49,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101}],"score":69.39,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":77450},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1650},"family":"GDP.pdf","row_ids":["792f8618c792ce48ee9c7c6155488bb6"],"score":86.84},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3550},"family":"AA-Omniscience","row_ids":["28a0e72d64e34dbdaef952ff1ff713b4","1a5d81b92d3a8437744af4b44b483171"],"score":66.98}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385}],"score":76.91,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1250},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1250},"family":"MMMU-Pro","row_ids":["e681be1aa8a9a82be96c894925b92cdd"],"score":73.53}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185}],"score":73.53,"weight":10}],"eligible":true,"id":"m-f33f6c88c27b2b9a","in_window":true,"name":"Inkling 2 Small","official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-11-05","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":15,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7ead6883bce4c045cf26f5b116b4cd48","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"0.1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0571428571428571,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0571428571428571},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"89d939473fcae282b052740fc3fcfd7a","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"Diamond","peer_count":64,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.287,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.287},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"34dffe7c2a7d437da340dfb5246fe9c8","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2 Hard","peer_count":52,"percentile":68.63,"percentile_exact":{"denominator":51,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.35,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"66284fe19cdf80bb97a995a814e23753","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":80.77,"percentile_exact":{"denominator":13,"numerator":1050},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.404123711340206,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.404123711340206},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f53e9f08adf2987f0e2c59d35bf57232","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v2.1","peer_count":82,"percentile":79.01,"percentile_exact":{"denominator":81,"numerator":6400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.23,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.23},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b334bad67fbd4a96eed8e669e0111fbb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"4.0","peer_count":76,"percentile":61.33,"percentile_exact":{"denominator":3,"numerator":184},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.131313131313131,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131313131313131},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb9e5b876d81fca5e8cf53e8169709cb","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.45621320178972646,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45621320178972646},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae0bf141fcba5d508b0da173dcf01538","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":78,"percentile":81.82,"percentile_exact":{"denominator":11,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1526.86,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1526.86},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ccd3047c248ced9697a972d3c394d08d","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA run","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.468218442256043,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.468218442256043},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e23e94bce569e3668ea49a1736f85e07","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA pass^5","peer_count":24,"percentile":76.09,"percentile_exact":{"denominator":23,"numerator":1750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"0fe380f78c53ebf4d268a83df481e0ad","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"2.0","peer_count":50,"percentile":69.39,"percentile_exact":{"denominator":49,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.25352601010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25352601010101},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"792f8618c792ce48ee9c7c6155488bb6","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.252,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.252},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28a0e72d64e34dbdaef952ff1ff713b4","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.33766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.33766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling 2 Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1a5d81b92d3a8437744af4b44b483171","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"v1.1","peer_count":54,"percentile":81.13,"percentile_exact":{"denominator":53,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7332662304982385,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7332662304982385},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling 2 Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e681be1aa8a9a82be96c894925b92cdd","included":true,"model_id":"m-f33f6c88c27b2b9a","original_version_label":"current Standard 10","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.815606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.815606936416185}],"release_date":"2026-11-05","release_status":"published","release_verified":true,"result_updated_at":null,"score":71.98,"score_exact":{"denominator":893888511516,"numerator":64345620454681},"source_id":"published/aa:inkling-small-1102-rc17"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":821457,"numerator":59247200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5350},"family":"GPQA Diamond","row_ids":["65e0dc8726d7a3e7afde224768d43963"],"score":90.68},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":950},"family":"Terminal-Bench-Science","row_ids":["4ea60fe9901ae1fc5d3c1daf559ab6e1"],"score":73.08},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4400},"family":"HLE","row_ids":["2ab50d9a5a2d86ccb853771d429875d3"],"score":69.84},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["b1d4b1cfdd4d38290545201e6c39db63"],"score":54.9}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","included":true,"model_id":"m-825009614c016c1d","original_version_label":"0.1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond","peer_count":64,"percentile":69.84,"percentile_exact":{"denominator":63,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":72.12,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":2267673408,"numerator":186901243525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["ae1b88ab2d0880e90cfd75002e9f4d77"],"score":100.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1225},"family":"tau3-Banking","row_ids":["1cd94c3f93e80676324cc2282ca34aa4"],"score":94.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7100},"family":"GDPval-AA","row_ids":["8fccbdc41b807b2ebbe41cff024b0690"],"score":87.65},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1675},"family":"Terminal-Bench","row_ids":["0d2e487f06bf23cbe7ee7f90f6003d16"],"score":69.79},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1275},"family":"AutomationBench","row_ids":["308f05669cd695ca55d1d348f97d3b83"],"score":67.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6900},"family":"AA-Briefcase","row_ids":["ecc6d7d48519d5a999fc5f71c76c0b6d"],"score":89.61},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":850},"family":"EnterpriseOps-Gym","row_ids":["aaa4039880a2e3c56a0fee65a0cf496e"],"score":94.44},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1300},"family":"AA-AnalystAgent","row_ids":["cee5a16789123af6ef334475ebff73a7"],"score":56.52}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","included":true,"model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":94.23,"percentile_exact":{"denominator":13,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2.1","peer_count":82,"percentile":87.65,"percentile_exact":{"denominator":81,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","included":true,"model_id":"m-825009614c016c1d","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","included":true,"model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":67.11,"percentile_exact":{"denominator":19,"numerator":1275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":78,"percentile":89.61,"percentile_exact":{"denominator":77,"numerator":6900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","peer_count":24,"percentile":56.52,"percentile_exact":{"denominator":23,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45}],"score":82.42,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":40975},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3100},"family":"FrontierCode","row_ids":["6cbe0b4839fdc145889ebffe2def2a50"],"score":63.27},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":850},"family":"Vals Code Migration","row_ids":["62c5b15d1c0217ab796423502244a099"],"score":65.38}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","included":true,"model_id":"m-825009614c016c1d","original_version_label":"2.0","peer_count":50,"percentile":63.27,"percentile_exact":{"denominator":49,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","included":true,"model_id":"m-825009614c016c1d","original_version_label":"1.1","peer_count":27,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19}],"score":64.32,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":2127575},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3225},"family":"GDP.pdf","row_ids":["5928f02dfd05733dd4d23f64991980e9"],"score":84.87},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":700},"family":"SimpleQA Verified","row_ids":["dd26bef806a0bba267bbd1b02770b43f"],"score":41.18},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3200},"family":"AA-Omniscience","row_ids":["c42d799db980160b63c3bab31d80fa02","602505b7f91e3456e6da6058f7b120bf"],"score":60.38}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","included":true,"model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088}],"score":62.14,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1300},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1300},"family":"MMMU-Pro","row_ids":["502735168b948696e5ae04c6357e58b6"],"score":76.47}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","peer_count":35,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884}],"score":76.47,"weight":10}],"eligible":true,"id":"m-825009614c016c1d","in_window":true,"name":"Qwen3.8 Max (0902)","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":16,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"65e0dc8726d7a3e7afde224768d43963","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-max:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ea60fe9901ae1fc5d3c1daf559ab6e1","included":true,"model_id":"m-825009614c016c1d","original_version_label":"0.1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.119047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-max:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.119047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2ab50d9a5a2d86ccb853771d429875d3","included":true,"model_id":"m-825009614c016c1d","original_version_label":"Diamond","peer_count":64,"percentile":69.84,"percentile_exact":{"denominator":63,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.24,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-max:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b1d4b1cfdd4d38290545201e6c39db63","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-max:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae1b88ab2d0880e90cfd75002e9f4d77","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.424041297935103,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-8-max:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.424041297935103},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1cd94c3f93e80676324cc2282ca34aa4","included":true,"model_id":"m-825009614c016c1d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":94.23,"percentile_exact":{"denominator":13,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.478350515463918,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-max:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.478350515463918},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8fccbdc41b807b2ebbe41cff024b0690","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v2.1","peer_count":82,"percentile":87.65,"percentile_exact":{"denominator":81,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1673.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-max:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1673.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"0d2e487f06bf23cbe7ee7f90f6003d16","included":true,"model_id":"m-825009614c016c1d","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row16","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"308f05669cd695ca55d1d348f97d3b83","included":true,"model_id":"m-825009614c016c1d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":67.11,"percentile_exact":{"denominator":19,"numerator":1275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.561905381087678,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-max:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.561905381087678},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ecc6d7d48519d5a999fc5f71c76c0b6d","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":78,"percentile":89.61,"percentile_exact":{"denominator":77,"numerator":6900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1606.14,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-max:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1606.14},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"aaa4039880a2e3c56a0fee65a0cf496e","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA run","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4762757385854969,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-max:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4762757385854969},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cee5a16789123af6ef334475ebff73a7","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA pass^5","peer_count":24,"percentile":56.52,"percentile_exact":{"denominator":23,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.45,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"qwen3-8-max:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6cbe0b4839fdc145889ebffe2def2a50","included":true,"model_id":"m-825009614c016c1d","original_version_label":"2.0","peer_count":50,"percentile":63.27,"percentile_exact":{"denominator":49,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.234232323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-max:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.234232323232323},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902) / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"62c5b15d1c0217ab796423502244a099","included":true,"model_id":"m-825009614c016c1d","original_version_label":"1.1","peer_count":27,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":57.19,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"57.19%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row10","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":57.19},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5928f02dfd05733dd4d23f64991980e9","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-max:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max-0902_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd26bef806a0bba267bbd1b02770b43f","included":true,"model_id":"m-825009614c016c1d","original_version_label":"run 1.2.0","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.473,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T18:04:58.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2ed9560be390bfe4db9b23c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"2ed9560be390bfe4db9b23c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.473},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c42d799db980160b63c3bab31d80fa02","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.29533333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.29533333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 Max (0902)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"602505b7f91e3456e6da6058f7b120bf","included":true,"model_id":"m-825009614c016c1d","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7294228949858088,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-max:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7294228949858088},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 Max (0902)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"502735168b948696e5ae04c6357e58b6","included":true,"model_id":"m-825009614c016c1d","original_version_label":"current Standard 10","peer_count":35,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.827745664739884,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-max:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-max:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.827745664739884}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"result_updated_at":"2026-09-02","score":71.58,"score_exact":{"denominator":3375323019484416,"numerator":241593097629493205},"source_id":"alibaba/qwen3.8-max-0902"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":3,"exact":{"denominator":2128,"numerator":133225},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":300},"family":"FrontierMath","row_ids":["88ff5cf09a8ea0336d12c5e7aac222d3"],"score":42.86},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":225},"family":"GPQA Diamond","row_ids":["e80c9d5d961020a2775a86f5a8e8c7ff"],"score":56.25},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1450},"family":"Terminal-Bench-Science","row_ids":["12c1a7b1c4eb18f0409ef269987fa5ca"],"score":76.32},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":75},"family":"HLE","row_ids":["852fa04b612ed916ac880c4c5a9b046b"],"score":75.0}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v2","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","peer_count":39,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6}],"score":62.61,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":6,"exact":{"denominator":252,"numerator":18965},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["803b3c7ebd5d583be94f718268d22790"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"GDPval-AA","row_ids":["3e2aa560fb85d043c269e4880c8d1ed7"],"score":66.67},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":325},"family":"Terminal-Bench","row_ids":["7df1cc76a5ce426695809152ab7c2b62"],"score":81.25},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"AutomationBench","row_ids":["ae96d3d1f5d89c47f0a27de7f5eca5d4"],"score":40.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":800},"family":"AA-Briefcase","row_ids":["a01ae7f3c128573c113df7e946ef1073"],"score":88.89},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":50},"family":"OSWorld","row_ids":["5f71dd3fff3f2fe7b85716b5a982127d"],"score":50.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":100},"family":"MCP Atlas","row_ids":["2ccdf17c2a9e8d9727ba57cd7a67ff5d"],"score":100.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","included":true,"model_id":"m-568e0eb3a95b5113","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","peer_count":49,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":7,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8}],"score":75.26,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":75},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":75},"family":"DeepSWE","row_ids":["d36cdba4dae45a238d9898e2ce34c241"],"score":75.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7}],"score":75.0,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":2,"exact":{"denominator":27,"numerator":1730},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1300},"family":"Humanity Sixth Sense","row_ids":["175d0f5744557d89454896c1420ebf13"],"score":48.15},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":80},"family":"Chartography","row_ids":["1ce2cb7b64555f9b41b3b5999f8e70ba"],"score":80.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","peer_count":28,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4}],"score":64.07,"weight":10}],"eligible":true,"id":"m-568e0eb3a95b5113","in_window":true,"name":"Claude Opus 5","official_evidence":{"artifact_sha256":"2f2269c3e1f9030bf9adc7e06e1979500968e3fc300eedd57a2b54424cab1694","basis":"official_release_page","date":"2026-07-24","sha256":"2f2269c3e1f9030bf9adc7e06e1979500968e3fc300eedd57a2b54424cab1694","url":"https://www.anthropic.com/news/claude-opus-5"},"primary_profile_id":"native","profile_id":"native","rank":17,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"88ff5cf09a8ea0336d12c5e7aac222d3","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v2","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":73.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"reasoning","direction":"higher_better","id":"e80c9d5d961020a2775a86f5a8e8c7ff","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":93.7,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"93.7%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":93.7},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"12c1a7b1c4eb18f0409ef269987fa5ca","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"0.1.0","peer_count":39,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":27.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"27.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":27.14},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"claude-opus-5","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"852fa04b612ed916ac880c4c5a9b046b","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"Diamond","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":38.6,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bce964f9a8360a45a7982916"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"bce964f9a8360a45a7982916","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":38.6},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Claude Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"803b3c7ebd5d583be94f718268d22790","included":true,"model_id":"m-568e0eb3a95b5113","peer_count":8,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":41.8,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"41.8","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Claude Opus 5"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":41.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"3e2aa560fb85d043c269e4880c8d1ed7","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1708.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1708","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1708.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7df1cc76a5ce426695809152ab7c2b62","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"4.0","peer_count":49,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":53.53,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"53.53%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row10"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row10","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":53.53},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"ae96d3d1f5d89c47f0a27de7f5eca5d4","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published developer protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":26.9,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"26.9%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column3","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":26.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 5 (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a01ae7f3c128573c113df7e946ef1073","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"v1.1","peer_count":19,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.48484848484848486,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ee2814c4220a7e6af2f98735"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ee2814c4220a7e6af2f98735","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.48484848484848486},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"5f71dd3fff3f2fe7b85716b5a982127d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"2.1","peer_count":7,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":74.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"74.0% partial","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row16:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-0dd3bed5c7dd:table0:row16:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":74.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-5 (xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"2ccdf17c2a9e8d9727ba57cd7a67ff5d","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":85.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"641cb1b24ad7b60c33181fc4"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"641cb1b24ad7b60c33181fc4","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":85.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 5","dimension":"coding","direction":"higher_better","id":"d36cdba4dae45a238d9898e2ce34c241","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"1.1","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column5","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"175d0f5744557d89454896c1420ebf13","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"current public protocol","peer_count":28,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cea8d8cb472002c4dccf0f7b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cea8d8cb472002c4dccf0f7b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.5},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"Opus 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"1ce2cb7b64555f9b41b3b5999f8e70ba","included":true,"model_id":"m-568e0eb3a95b5113","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":83.4,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"83.4% with tools","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row18:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-0dd3bed5c7dd:table0:row18:column3","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"%","value":83.4}],"release_date":"2026-07-24","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":69.84,"score_exact":{"denominator":131328,"numerator":9172165},"source_id":"anthropic/claude-opus-5"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":61047,"numerator":3863750},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["5f766555b014c71ff2d651ccb2fc76d7"],"score":57.89},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4300},"family":"HLE","row_ids":["353c470ec7dd55d7fbbc616285180c31"],"score":68.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3250},"family":"CritPt","row_ids":["539eef88fa72b976d41f8f0e6b20d640"],"score":63.73}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","peer_count":64,"percentile":68.25,"percentile_exact":{"denominator":63,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335}],"score":63.29,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2725569,"numerator":228484706},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7300},"family":"GDPval-AA","row_ids":["4d4ab3fcd8c470a4ca605a1e507ade3e"],"score":90.12},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":70},"family":"Terminal-Bench","row_ids":["a073820ddc97f09d7966e361779cf68e"],"score":70.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1600},"family":"AutomationBench","row_ids":["5dc195401c512a05556701293da42306"],"score":84.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":7100},"family":"AA-Briefcase","row_ids":["7ce9f4dd8dbbce093592c7a8603e6868"],"score":92.21},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1900},"family":"AA-AnalystAgent","row_ids":["1c70059543aa3443966891f5d87280d7"],"score":82.61}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","peer_count":82,"percentile":90.12,"percentile_exact":{"denominator":81,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":84.21,"percentile_exact":{"denominator":19,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":78,"percentile":92.21,"percentile_exact":{"denominator":77,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","peer_count":24,"percentile":82.61,"percentile_exact":{"denominator":23,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525}],"score":83.83,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":43725},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3900},"family":"FrontierCode","row_ids":["efab926973de754a7bc887e6a5732e79"],"score":79.59},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":750},"family":"Vals Code Migration","row_ids":["575f9b1e7b6ccadadc3e13c6241b21d4"],"score":57.69}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"2.0","peer_count":50,"percentile":79.59,"percentile_exact":{"denominator":49,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"1.1","peer_count":27,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22}],"score":68.64,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":901,"numerator":65725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"GDP.pdf","row_ids":["06f207d047d6194d347719cbe681b11e"],"score":75.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1050},"family":"SimpleQA Verified","row_ids":["1bbbdc3f7be54b8c3de3afd5eb04f4c5"],"score":61.76},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":4350},"family":"AA-Omniscience","row_ids":["86bdedd227d21984b4d30c6ee2bdc8b4","5789f26e6b16a0b75b183dcbb8cf44bd"],"score":82.08}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947}],"score":72.95,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":700},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":700},"family":"MMMU-Pro","row_ids":["6a15ad1c8de4bbfb578c4f789e53c9b8"],"score":41.18}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341}],"score":41.18,"weight":10}],"eligible":true,"id":"m-7187cf2b01d41935","in_window":true,"name":"Grok 4.7","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":18,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.7 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5f766555b014c71ff2d651ccb2fc76d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"353c470ec7dd55d7fbbc616285180c31","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"Diamond","peer_count":64,"percentile":68.25,"percentile_exact":{"denominator":63,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.232,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"grok-4-7:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"539eef88fa72b976d41f8f0e6b20d640","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"grok-4-7:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4d4ab3fcd8c470a4ca605a1e507ade3e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v2.1","peer_count":82,"percentile":90.12,"percentile_exact":{"denominator":81,"numerator":7300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1718.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"grok-4-7:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1718.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a073820ddc97f09d7966e361779cf68e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"grok-4-7:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5dc195401c512a05556701293da42306","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":84.21,"percentile_exact":{"denominator":19,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6555835148969458,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"grok-4-7:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6555835148969458},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ce9f4dd8dbbce093592c7a8603e6868","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":78,"percentile":92.21,"percentile_exact":{"denominator":77,"numerator":7100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1637.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"grok-4-7:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1637.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1c70059543aa3443966891f5d87280d7","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA pass^5","peer_count":24,"percentile":82.61,"percentile_exact":{"denominator":23,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.525,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"grok-4-7:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.525},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"efab926973de754a7bc887e6a5732e79","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"2.0","peer_count":50,"percentile":79.59,"percentile_exact":{"denominator":49,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.286575505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"grok-4-7:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.286575505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Grok 4.7 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"575f9b1e7b6ccadadc3e13c6241b21d4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"1.1","peer_count":27,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":53.22,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"53.22%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row12","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":53.22},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"06f207d047d6194d347719cbe681b11e","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current AA protocol","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.2,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"grok-4-7:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.7_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"1bbbdc3f7be54b8c3de3afd5eb04f4c5","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"run 1.2.0","peer_count":35,"percentile":61.76,"percentile_exact":{"denominator":17,"numerator":1050},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.56,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T21:49:22.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fc8ef5e86f988f57cc6e2a6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0fc8ef5e86f988f57cc6e2a6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.56},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"86bdedd227d21984b4d30c6ee2bdc8b4","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.427,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.427},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Grok 4.7 (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5789f26e6b16a0b75b183dcbb8cf44bd","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"v1.1","peer_count":54,"percentile":90.57,"percentile_exact":{"denominator":53,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7998836532867947,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"grok-4-7:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7998836532867947},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Grok 4.7 (Low)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"6a15ad1c8de4bbfb578c4f789e53c9b8","included":true,"model_id":"m-7187cf2b01d41935","original_version_label":"current Standard 10","peer_count":35,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.78150289017341,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"grok-4-7-low:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"grok-4-7-low:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.78150289017341}],"release_date":"2026-09-21","release_status":"published","release_verified":true,"result_updated_at":"2026-09-22","score":69.22,"score_exact":{"denominator":223472127879,"numerator":15467797601309},"source_id":"xai/grok-4.7"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":114342,"numerator":7985875},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5800},"family":"GPQA Diamond","row_ids":["93a25d2afe298e1250966bfe70f4f70f"],"score":98.31},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1000},"family":"Terminal-Bench-Science","row_ids":["4c2b94e36c85b9d8f2e3d83490a34034"],"score":52.63},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":50},"family":"HLE","row_ids":["08dc502d8d4aeb42de93130ed556d391"],"score":50.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":4000},"family":"CritPt","row_ids":["4e088fc1433335f9f212c9f9ffa8b974"],"score":78.43}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":98.31,"percentile_exact":{"denominator":59,"numerator":5800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","peer_count":39,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","peer_count":17,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","peer_count":52,"percentile":78.43,"percentile_exact":{"denominator":51,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465}],"score":69.84,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":5594589,"numerator":380137666},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1125},"family":"tau3-Banking","row_ids":["ceb2e04d7badea9a2869bdbce0586635"],"score":86.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5300},"family":"GDPval-AA","row_ids":["5e4d6ed857d58561b430218e5d2e3d99"],"score":65.43},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":64},"family":"Terminal-Bench","row_ids":["477576b238b66114c394a26ca4a20345"],"score":64.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"AutomationBench","row_ids":["44e1f8cce1da50e102dced9197ed9f4e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4300},"family":"AA-Briefcase","row_ids":["c6c37db45460e80aac980c7255d8ffed"],"score":55.84},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1400},"family":"AA-AnalystAgent","row_ids":["9cd6a63a2e9021c4431ddd906f3978b5"],"score":60.87}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":86.54,"percentile_exact":{"denominator":13,"numerator":1125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","peer_count":82,"percentile":65.43,"percentile_exact":{"denominator":81,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"4.0","peer_count":76,"percentile":64.0,"percentile_exact":{"denominator":1,"numerator":64},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":78,"percentile":55.84,"percentile_exact":{"denominator":77,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","peer_count":24,"percentile":60.87,"percentile_exact":{"denominator":23,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625}],"score":67.95,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":3822,"numerator":208475},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2600},"family":"FrontierCode","row_ids":["926e007cfcb9b2d7a75e59641d96bd9e"],"score":53.06},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":300},"family":"Vals Code Migration","row_ids":["59c32ec585e3e137a1a5f19c3bda0e98"],"score":23.08},{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":175},"family":"DeepSWE","row_ids":["e589efcb9de8d6dea5758589d780110f"],"score":87.5}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"2.0","peer_count":50,"percentile":53.06,"percentile_exact":{"denominator":49,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":27,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":9,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8}],"score":54.55,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":3835475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1525},"family":"GDP.pdf","row_ids":["1bf434a1cea1439ab199e26911f14113"],"score":80.26},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1450},"family":"SimpleQA Verified","row_ids":["f06072893cbefd23dc63a86432677691"],"score":85.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3100},"family":"AA-Omniscience","row_ids":["1cb384dc9fec91c5db3c1841d018dfc3","4b5e9db335123a5e12d2351e281e6182"],"score":58.49}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916}],"score":74.68,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":38950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2200},"family":"Humanity Sixth Sense","row_ids":["11851315d0cbcd3709f1ea2206b9b74b"],"score":81.48},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1500},"family":"MMMU-Pro","row_ids":["8bcd59ede7a9782bc9b8b3c169c19887"],"score":88.24}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","peer_count":28,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185}],"score":84.86,"weight":10}],"eligible":true,"id":"m-422e5ba87859dd55","in_window":true,"name":"Gemini 3.8 Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":19,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"93a25d2afe298e1250966bfe70f4f70f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":98.31,"percentile_exact":{"denominator":59,"numerator":5800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.952525252525253,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-8-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.952525252525253},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"4c2b94e36c85b9d8f2e3d83490a34034","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"0.1.0","peer_count":39,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.57,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"8.57%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":8.57},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gemini-3.8-flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"08dc502d8d4aeb42de93130ed556d391","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"Diamond","peer_count":17,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":34.3,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"b6dc9eb1bbec2999597adc1d"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"b6dc9eb1bbec2999597adc1d","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":34.3},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4e088fc1433335f9f212c9f9ffa8b974","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2 Hard","peer_count":52,"percentile":78.43,"percentile_exact":{"denominator":51,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.465,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-8-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.465},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ceb2e04d7badea9a2869bdbce0586635","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":86.54,"percentile_exact":{"denominator":13,"numerator":1125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.449484536082474,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-8-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.449484536082474},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5e4d6ed857d58561b430218e5d2e3d99","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v2.1","peer_count":82,"percentile":65.43,"percentile_exact":{"denominator":81,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1445.43,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-8-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1445.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"477576b238b66114c394a26ca4a20345","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"4.0","peer_count":76,"percentile":64.0,"percentile_exact":{"denominator":1,"numerator":64},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.196969696969697,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-8-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.196969696969697},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"44e1f8cce1da50e102dced9197ed9f4e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5993009432118243,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-8-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5993009432118243},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6c37db45460e80aac980c7255d8ffed","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":78,"percentile":55.84,"percentile_exact":{"denominator":77,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1202.99,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-8-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1202.99},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9cd6a63a2e9021c4431ddd906f3978b5","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA pass^5","peer_count":24,"percentile":60.87,"percentile_exact":{"denominator":23,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gemini-3-8-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"926e007cfcb9b2d7a75e59641d96bd9e","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"2.0","peer_count":50,"percentile":53.06,"percentile_exact":{"denominator":49,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.187427777777778,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-8-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187427777777778},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Gemini 3.8 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"59c32ec585e3e137a1a5f19c3bda0e98","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":27,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":38.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"38.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row22","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":38.65},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.8 Flash","dimension":"coding","direction":"higher_better","id":"e589efcb9de8d6dea5758589d780110f","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"1.1","peer_count":9,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":73.8,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column6","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":73.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1bf434a1cea1439ab199e26911f14113","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current AA protocol","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.21,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-8-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.8-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f06072893cbefd23dc63a86432677691","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"run 1.2.0","peer_count":35,"percentile":85.29,"percentile_exact":{"denominator":17,"numerator":1450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.697,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-02","result_timestamp":"2026-09-02T17:15:36.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9db450757aab5ef9e099a74b"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9db450757aab5ef9e099a74b","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.697},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1cb384dc9fec91c5db3c1841d018dfc3","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":88.68,"percentile_exact":{"denominator":53,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5053333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5053333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.8 Flash (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4b5e9db335123a5e12d2351e281e6182","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"v1.1","peer_count":54,"percentile":28.3,"percentile_exact":{"denominator":53,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44878706199460916,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-8-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44878706199460916},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.8-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"11851315d0cbcd3709f1ea2206b9b74b","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current public protocol","peer_count":28,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":41.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"76ddad5bf102925e95027213"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"76ddad5bf102925e95027213","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":41.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.8 Flash (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"8bcd59ede7a9782bc9b8b3c169c19887","included":true,"model_id":"m-422e5ba87859dd55","original_version_label":"current Standard 10","peer_count":35,"percentile":88.24,"percentile_exact":{"denominator":17,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.85606936416185,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-8-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-8-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.85606936416185}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":68.78,"score_exact":{"denominator":18613913710392,"numerator":1280244857994287},"source_id":"google/gemini-3.8-flash"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":41.67,"evaluator_count":1,"exact":{"denominator":59,"numerator":4750},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4750},"family":"GPQA Diamond","row_ids":["2f5e3b6ebf2112f4f734c9c632e248bf"],"score":80.51}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505}],"score":80.51,"weight":25},{"effective_weight_percent":41.67,"evaluator_count":1,"exact":{"denominator":54,"numerator":3625},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":75},"family":"tau3-Banking","row_ids":["569f96b9f64d6f89c3e721fb138a9646"],"score":75.0},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1600},"family":"GDPval-AA","row_ids":["67407c66c8ccc8ec87ccd4b6ae121aa3"],"score":59.26}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"v2.1","peer_count":82,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33}],"score":67.13,"weight":25},{"effective_weight_percent":16.67,"evaluator_count":1,"exact":{"denominator":17,"numerator":550},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":550},"family":"MMMU-Pro","row_ids":["d7ea5990ff16dd400cf1a637a1392f4c"],"score":32.35}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607}],"score":32.35,"weight":10}],"eligible":true,"id":"m-e060cf24e4176269","in_window":true,"name":"Agnes 2.5 Pro Beta","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":20,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f5e3b6ebf2112f4f734c9c632e248bf","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"agnes-2-5-pro-beta:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"569f96b9f64d6f89c3e721fb138a9646","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.356701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"agnes-2-5-pro-beta:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.356701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"67407c66c8ccc8ec87ccd4b6ae121aa3","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"v2.1","peer_count":82,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.33,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"agnes-2-5-pro-beta:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Agnes 2.5 Pro Beta","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d7ea5990ff16dd400cf1a637a1392f4c","included":true,"model_id":"m-e060cf24e4176269","original_version_label":"current Standard 10","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.754335260115607,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"agnes-2-5-pro-beta:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"agnes-2-5-pro-beta:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.754335260115607}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"result_updated_at":null,"score":66.91,"score_exact":{"denominator":649944,"numerator":43486475},"source_id":"published/aa:agnes-2-5-pro-beta"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":37128,"numerator":2162375},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":350},"family":"FrontierMath","row_ids":["f2e8485b9221333706b0e9e14cd8c4e7"],"score":38.89},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1975},"family":"Terminal-Bench-Science","row_ids":["ebff4c97ce1c3d30b95d6d855419e5c1"],"score":75.96},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4600},"family":"HLE","row_ids":["2bfa42798a174eb5a88b71f065cd4aef"],"score":73.02},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2300},"family":"CritPt","row_ids":["d9b68c289d550eec511fa758a53ee05a"],"score":45.1}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","peer_count":64,"percentile":73.02,"percentile_exact":{"denominator":63,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","peer_count":52,"percentile":45.1,"percentile_exact":{"denominator":51,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25}],"score":58.24,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114912,"numerator":9263075},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2500},"family":"GDPval-AA","row_ids":["0a034403545d9a6f875f18bdfdb8614b"],"score":92.59},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1375},"family":"Terminal-Bench","row_ids":["6d06cabd5082651a9ea6ff6f7709c136"],"score":57.29},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1650},"family":"AutomationBench","row_ids":["74476994a8c12c8ddef6748c439b2908"],"score":86.84},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":600},"family":"AA-Briefcase","row_ids":["e9014188fe35aeb589dfba302eb5ab58"],"score":85.71}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","peer_count":82,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":78,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42}],"score":80.61,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":49200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3800},"family":"FrontierCode","row_ids":["981d7322c13b53f8d5edfa1fdb6e95e5"],"score":77.55},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":1000},"family":"Vals Code Migration","row_ids":["1080a38736b6ba3768e68ace6a340f30"],"score":76.92}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","peer_count":50,"percentile":77.55,"percentile_exact":{"denominator":49,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","peer_count":27,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52}],"score":77.24,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2456375},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1325},"family":"GDP.pdf","row_ids":["b06f9a392143fecd0d26f4f434ebd403"],"score":69.74},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":500},"family":"SimpleQA Verified","row_ids":["8e35d6850446c900d7c667bcebf2d5eb"],"score":29.41},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2350},"family":"AA-Omniscience","row_ids":["28843797b0f403bc00b3b129a0250f86","cd12ad2957c69fc5e8660a959eaae5db"],"score":44.34}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795}],"score":47.83,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":153,"numerator":8575},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"Humanity Sixth Sense","row_ids":["60e8851ddeb919672b146036d9d51182"],"score":44.44},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1150},"family":"MMMU-Pro","row_ids":["33026de8036ee6188f530a53e9aaa0aa"],"score":67.65}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","peer_count":28,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289}],"score":56.05,"weight":10}],"eligible":true,"id":"m-caf1ef1f3bdf5452","in_window":true,"name":"DeepSeek V4.1 Flash 1017","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-17","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":21,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f2e8485b9221333706b0e9e14cd8c4e7","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:39:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2279d24f52a49e406317ac69"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"2279d24f52a49e406317ac69","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ebff4c97ce1c3d30b95d6d855419e5c1","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"0.1","peer_count":53,"percentile":75.96,"percentile_exact":{"denominator":26,"numerator":1975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.128571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"deepseek-v4-1-flash-1017:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2bfa42798a174eb5a88b71f065cd4aef","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"Diamond","peer_count":64,"percentile":73.02,"percentile_exact":{"denominator":63,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.253,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-1-flash-1017:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.253},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d9b68c289d550eec511fa758a53ee05a","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2 Hard","peer_count":52,"percentile":45.1,"percentile_exact":{"denominator":51,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.25,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-1-flash-1017:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0a034403545d9a6f875f18bdfdb8614b","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v2.1","peer_count":82,"percentile":92.59,"percentile_exact":{"denominator":27,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1745.77,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-1-flash-1017:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1745.77},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6d06cabd5082651a9ea6ff6f7709c136","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"74476994a8c12c8ddef6748c439b2908","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":86.84,"percentile_exact":{"denominator":19,"numerator":1650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6664095364949257,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-1-flash-1017:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6664095364949257},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9014188fe35aeb589dfba302eb5ab58","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":78,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1570.42,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-1-flash-1017:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1570.42},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"981d7322c13b53f8d5edfa1fdb6e95e5","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"2.0","peer_count":50,"percentile":77.55,"percentile_exact":{"denominator":49,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.284711616161616,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-1-flash-1017:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.284711616161616},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1080a38736b6ba3768e68ace6a340f30","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"1.1","peer_count":27,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":67.52,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"67.52%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row7"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row7","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":67.52},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b06f9a392143fecd0d26f4f434ebd403","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current AA protocol","peer_count":77,"percentile":69.74,"percentile_exact":{"denominator":19,"numerator":1325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.172,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-1-flash-1017:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash-1017_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8e35d6850446c900d7c667bcebf2d5eb","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"run 1.2.0","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.425,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-19","result_timestamp":"2026-10-19T22:38:54.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dd818b53eb4225f783b4bd21"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dd818b53eb4225f783b4bd21","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.425},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"28843797b0f403bc00b3b129a0250f86","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.396,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.396},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd12ad2957c69fc5e8660a959eaae5db","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.35927152317880795,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-1-flash-1017:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35927152317880795},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"60e8851ddeb919672b146036d9d51182","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current public protocol","peer_count":28,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30.1,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"57a53b70701c1a5254ae7d9f"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"57a53b70701c1a5254ae7d9f","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30.1},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash 1017 (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"33026de8036ee6188f530a53e9aaa0aa","included":true,"model_id":"m-caf1ef1f3bdf5452","original_version_label":"current Standard 10","peer_count":35,"percentile":67.65,"percentile_exact":{"denominator":17,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.80635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-1-flash-1017:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-1-flash-1017:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.80635838150289}],"release_date":"2026-10-17","release_status":"published","release_verified":true,"result_updated_at":"2026-10-19","score":65.33,"score_exact":{"denominator":37686999168,"numerator":2462117234705},"source_id":"deepseek/deepseek-v4.1-flash-1017"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":152,"numerator":9325},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1500},"family":"Terminal-Bench-Science","row_ids":["f1163968f642895b63614f184b419fdc"],"score":78.95},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":175},"family":"HLE","row_ids":["3ff9ef5159d8b127a6433bdd9ec141a5"],"score":43.75}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","peer_count":39,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8}],"score":61.35,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":1235},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":200},"family":"GDPval-AA","row_ids":["8b095392eb68858c1d89a63205401933"],"score":22.22},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":75},"family":"Terminal-Bench","row_ids":["a142db33093dc7d6f056b8d8509f429e"],"score":75.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":40},"family":"AA-Briefcase","row_ids":["ad451031d1d4a9ddfab5c379bfc5d852"],"score":40.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","peer_count":49,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"reference_revision":"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0}],"score":45.74,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["0c33bab4ff416863111d8ddbb4eb7f0b"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8}],"score":100.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1250},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1250},"family":"SimpleQA Verified","row_ids":["8a0eac6d7e9eaac51a5eeee920a35f4d"],"score":73.53}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607}],"score":73.53,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":970},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1400},"family":"Humanity Sixth Sense","row_ids":["e25930e774ad7abbc0f5c7ffcc567aa5"],"score":51.85},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"Chartography","row_ids":["961e2599d1347e8ad57a29581acdb50c"],"score":20.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","peer_count":28,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6}],"score":35.93,"weight":10}],"eligible":true,"id":"m-3b5ea6fdfa94bc08","in_window":true,"name":"GPT-6 Sol","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":22,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f1163968f642895b63614f184b419fdc","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"0.1.0","peer_count":39,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":30.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"30.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row9","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":30.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"3ff9ef5159d8b127a6433bdd9ec141a5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"Diamond","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":33.8,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"859863b0a938dd900a6c81dc"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"859863b0a938dd900a6c81dc","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":33.8},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"8b095392eb68858c1d89a63205401933","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"2.1","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1487.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1487⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1487.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a142db33093dc7d6f056b8d8509f429e","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"4.0","peer_count":49,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":44.44,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"44.44%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":44.44},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"ad451031d1d4a9ddfab5c379bfc5d852","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work AA-Briefcase v1.1³; preserve page footnotes and task-specific settings","protocol_id":"74db141f961ec29fa7ab8d6a","raw_value":1483.0,"reference_revision":"e71abc51b541cf1e03f4dbbea41f9fa64ca4c089fe179314e77fab6ba6130dde","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1483⁴","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row11:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AA-Briefcase:Knowledge work AA-Briefcase v1.1","source_record_id":"official-4e3ca8a95d36:table0:row11:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":1483.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Sol (max)","dimension":"coding","direction":"higher_better","id":"0c33bab4ff416863111d8ddbb4eb7f0b","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"1.1","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":68.8,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Sol:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Sol:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":68.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-sol_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8a0eac6d7e9eaac51a5eeee920a35f4d","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"run 1.2.0","peer_count":35,"percentile":73.53,"percentile_exact":{"denominator":17,"numerator":1250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.607,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T19:59:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d5b3b2608201a4eea58f25f"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"7d5b3b2608201a4eea58f25f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.607},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e25930e774ad7abbc0f5c7ffcc567aa5","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"current public protocol","peer_count":28,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9151f6417f13ebba5e5c142b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9151f6417f13ebba5e5c142b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.2},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"GPT-6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"961e2599d1347e8ad57a29581acdb50c","included":true,"model_id":"m-3b5ea6fdfa94bc08","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":53.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"53.6%⁴ no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column4"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column4","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":53.6}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":"2026-10-26","score":65.07,"score_exact":{"denominator":93024,"numerator":6053149},"source_id":"openai/gpt-6-sol"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1071,"numerator":66400},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":50},"family":"Terminal-Bench-Science","row_ids":["e374e482dae499c526f6a7da9baef1aa"],"score":50.0},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1600},"family":"HLE","row_ids":["a1e07609863e9a13d3798af4a816cb48"],"score":76.19},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3050},"family":"CritPt","row_ids":["321a79f217d130c0ceb39102c3c1fab9"],"score":59.8}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"0.1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","peer_count":64,"percentile":76.19,"percentile_exact":{"denominator":21,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325}],"score":62.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":474012,"numerator":36496535},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5500},"family":"GDPval-AA","row_ids":["cfa7d9348c146c713f91a2534eae2b29"],"score":67.9},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":70},"family":"Terminal-Bench","row_ids":["3a8fd5af00baca1a19eb582e1a8c4ab5"],"score":70.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1825},"family":"AutomationBench","row_ids":["b04984f7d9488e18c36d59811ea3856b"],"score":96.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5700},"family":"AA-Briefcase","row_ids":["da58b11d089faf8c31c386386035abd7"],"score":74.03}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","peer_count":82,"percentile":67.9,"percentile_exact":{"denominator":81,"numerator":5500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":96.05,"percentile_exact":{"denominator":19,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":78,"percentile":74.03,"percentile_exact":{"denominator":77,"numerator":5700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98}],"score":76.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":2700},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2700},"family":"FrontierCode","row_ids":["1935649843d5e809cef231b07734be32"],"score":55.1}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"2.0","peer_count":50,"percentile":55.1,"percentile_exact":{"denominator":49,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182}],"score":55.1,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":237475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2975},"family":"GDP.pdf","row_ids":["96a35bebfd1cf232725317d8f36f4c53"],"score":78.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["2ab90fc472453676cb1037409906c906","cd0ef2b66ec6818e014beedc4ad88c9f"],"score":39.62}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483}],"score":58.96,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":900},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":900},"family":"MMMU-Pro","row_ids":["fc2b137ca09cf3cd0123e67faaeeed34"],"score":52.94}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786}],"score":52.94,"weight":10}],"eligible":true,"id":"m-8e76831a5814eec9","in_window":true,"name":"MiniMax-M3.1-Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-19","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":23,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e374e482dae499c526f6a7da9baef1aa","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"0.1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0333333333333333,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3-1-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0333333333333333},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a1e07609863e9a13d3798af4a816cb48","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"Diamond","peer_count":64,"percentile":76.19,"percentile_exact":{"denominator":21,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.281,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3-1-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.281},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"321a79f217d130c0ceb39102c3c1fab9","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3-1-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cfa7d9348c146c713f91a2534eae2b29","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v2.1","peer_count":82,"percentile":67.9,"percentile_exact":{"denominator":81,"numerator":5500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1465.36,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1465.36},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a8fd5af00baca1a19eb582e1a8c4ab5","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"4.0","peer_count":76,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.257575757575758,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minimax-m3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.257575757575758},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b04984f7d9488e18c36d59811ea3856b","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":96.05,"percentile_exact":{"denominator":19,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.7252180957532544,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7252180957532544},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da58b11d089faf8c31c386386035abd7","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":78,"percentile":74.03,"percentile_exact":{"denominator":77,"numerator":5700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1475.98,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1475.98},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"1935649843d5e809cef231b07734be32","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"2.0","peer_count":50,"percentile":55.1,"percentile_exact":{"denominator":49,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.208218181818182,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3-1-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208218181818182},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96a35bebfd1cf232725317d8f36f4c53","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2ab90fc472453676cb1037409906c906","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":77.36,"percentile_exact":{"denominator":53,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.45066666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.45066666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3.1-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"cd0ef2b66ec6818e014beedc4ad88c9f","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.06310679611650483,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3-1-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06310679611650483},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fc2b137ca09cf3cd0123e67faaeeed34","included":true,"model_id":"m-8e76831a5814eec9","original_version_label":"current Standard 10","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791329479768786,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3-1-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3-1-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791329479768786}],"release_date":"2026-10-19","release_status":"published","release_verified":true,"result_updated_at":null,"score":62.85,"score_exact":{"denominator":703433808,"numerator":44213635945},"source_id":"minimax/MiniMax-M3.1-Flash"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":3717,"numerator":278450},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5000},"family":"GPQA Diamond","row_ids":["185f07372e1def6a3525e6b1b6c635f5"],"score":84.75},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":4100},"family":"HLE","row_ids":["c42c9d1f70b755ec31ff5dcd6c984f87"],"score":65.08}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":84.75,"percentile_exact":{"denominator":59,"numerator":5000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","peer_count":64,"percentile":65.08,"percentile_exact":{"denominator":63,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211}],"score":74.91,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":105856906},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1075},"family":"tau3-Banking","row_ids":["b1886f41be6be37742b893448f0c7326"],"score":82.69},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1900},"family":"GDPval-AA","row_ids":["6a2341b888c3c4634de9d8d25fc69ba9"],"score":70.37},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":172},"family":"Terminal-Bench","row_ids":["cdfb425c1664f272548a144cf79359c8"],"score":57.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1075},"family":"AutomationBench","row_ids":["6ca0b1280bb8cb8f57834c52f2472139"],"score":56.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5200},"family":"AA-Briefcase","row_ids":["94b4404155c32f58241674a265734f95"],"score":67.53},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":700},"family":"EnterpriseOps-Gym","row_ids":["54f26e677fb3a678985d92f416cf28ad"],"score":77.78}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","peer_count":82,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"4.0","peer_count":76,"percentile":57.33,"percentile_exact":{"denominator":3,"numerator":172},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","peer_count":78,"percentile":67.53,"percentile_exact":{"denominator":77,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943}],"score":68.71,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":1050},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1050},"family":"GDP.pdf","row_ids":["b7c40d3349d6abfac4e0b93abf2adfa2"],"score":55.26}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12}],"score":55.26,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":500},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":500},"family":"MMMU-Pro","row_ids":["dbca8fa745f042a216711b700c5e06ca"],"score":29.41}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717}],"score":29.41,"weight":10}],"eligible":true,"id":"m-65378fc44ea2de26","in_window":true,"name":"DeepSeek V4 Flash Vision","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":24,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"185f07372e1def6a3525e6b1b6c635f5","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":84.75,"percentile_exact":{"denominator":59,"numerator":5000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.913131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-flash-vision:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.913131313131313},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c42c9d1f70b755ec31ff5dcd6c984f87","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"Diamond","peer_count":64,"percentile":65.08,"percentile_exact":{"denominator":63,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.211,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-flash-vision:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.211},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1886f41be6be37742b893448f0c7326","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":82.69,"percentile_exact":{"denominator":13,"numerator":1075},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.410309278350515,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-flash-vision:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.410309278350515},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6a2341b888c3c4634de9d8d25fc69ba9","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v2.1","peer_count":82,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1554.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-flash-vision:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1554.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cdfb425c1664f272548a144cf79359c8","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"4.0","peer_count":76,"percentile":57.33,"percentile_exact":{"denominator":3,"numerator":172},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.121212121212121,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"deepseek-v4-flash-vision:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121212121212121},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6ca0b1280bb8cb8f57834c52f2472139","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4747087040091967,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-flash-vision:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4747087040091967},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94b4404155c32f58241674a265734f95","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"v1.1","peer_count":78,"percentile":67.53,"percentile_exact":{"denominator":77,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1418.4,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-flash-vision:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1418.4},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54f26e677fb3a678985d92f416cf28ad","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA run","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4709042076991943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-flash-vision:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4709042076991943},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b7c40d3349d6abfac4e0b93abf2adfa2","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current AA protocol","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.12,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-flash-vision:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Flash Vision (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"dbca8fa745f042a216711b700c5e06ca","included":true,"model_id":"m-65378fc44ea2de26","original_version_label":"current Standard 10","peer_count":35,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.747976878612717,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-flash-vision:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"deepseek-v4-flash-vision:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.747976878612717}],"release_date":"2026-08-21","release_status":"published","release_verified":true,"result_updated_at":null,"score":62.38,"score_exact":{"denominator":3090321234,"numerator":192760715305},"source_id":"published/aa:deepseek-v4-flash-vision"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":5967,"numerator":358000},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["6b7d4617c648ad490ad35eb982d8d2e7"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":500},"family":"HLE","row_ids":["39803a6fda08687f289513743c809091"],"score":55.56},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1200},"family":"CritPt","row_ids":["20d35e57ebd245b7509c38232fe59998"],"score":70.59}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","peer_count":64,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","peer_count":52,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385}],"score":60.0,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":158004,"numerator":12977233},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2300},"family":"GDPval-AA","row_ids":["f04f5a2ca29d7ef8a19efd2486d1e259"],"score":85.19},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":224},"family":"Terminal-Bench","row_ids":["cf0c6cf90a55894b4c003eacdea2ccb7"],"score":74.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1675},"family":"AutomationBench","row_ids":["115a839b24d533001f549fec19f24cc9"],"score":88.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6200},"family":"AA-Briefcase","row_ids":["2d7e1c988db8f9f1d1a2378f0202de51"],"score":80.52}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","peer_count":82,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":78,"percentile":80.52,"percentile_exact":{"denominator":77,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522}],"score":82.13,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":2300},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2300},"family":"FrontierCode","row_ids":["a3fd831de30519528e597eb5e2ceea3e"],"score":46.94}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","peer_count":50,"percentile":46.94,"percentile_exact":{"denominator":49,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151}],"score":46.94,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":205325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2225},"family":"GDP.pdf","row_ids":["d7a0ab07d52625e0d345f1bbf275edda"],"score":58.55},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["568a15734298d304532a5f2a4188d632","ae54534276a6c4c19dcb828d96fdee4c"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247}],"score":50.97,"weight":20}],"eligible":true,"id":"m-a3f1f1de993a039e","in_window":true,"name":"Ling 3.1 Flash 1019","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":25,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6b7d4617c648ad490ad35eb982d8d2e7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"ling-3-1-flash-1019:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"39803a6fda08687f289513743c809091","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"Diamond","peer_count":64,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.172,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-1-flash-1019:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.172},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20d35e57ebd245b7509c38232fe59998","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2 Hard","peer_count":52,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.385,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"ling-3-1-flash-1019:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f04f5a2ca29d7ef8a19efd2486d1e259","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v2.1","peer_count":82,"percentile":85.19,"percentile_exact":{"denominator":27,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1652.94,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash-1019:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1652.94},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf0c6cf90a55894b4c003eacdea2ccb7","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash-1019:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"115a839b24d533001f549fec19f24cc9","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":88.16,"percentile_exact":{"denominator":19,"numerator":1675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6754660741517774,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash-1019:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6754660741517774},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d7e1c988db8f9f1d1a2378f0202de51","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":78,"percentile":80.52,"percentile_exact":{"denominator":77,"numerator":6200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1522,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash-1019:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1522},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a3fd831de30519528e597eb5e2ceea3e","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"2.0","peer_count":50,"percentile":46.94,"percentile_exact":{"denominator":49,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170090151515151,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"ling-3-1-flash-1019:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170090151515151},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7a0ab07d52625e0d345f1bbf275edda","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash-1019:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"568a15734298d304532a5f2a4188d632","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":35.85,"percentile_exact":{"denominator":53,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2653333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2653333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.1 Flash 1019","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ae54534276a6c4c19dcb828d96fdee4c","included":true,"model_id":"m-a3f1f1de993a039e","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6370235934664247,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash-1019:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-1-flash-1019:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6370235934664247}],"release_date":"2026-10-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":61.24,"score_exact":{"denominator":233188307352,"numerator":14280161662615},"source_id":"published/aa:ling-3-1-flash-rc5"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":1064,"numerator":64025},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":400},"family":"FrontierMath","row_ids":["2e3935359cea7e94134e77b2b6c16a83"],"score":57.14},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":75},"family":"GPQA Diamond","row_ids":["06022591029d9808b7b2c118e22277aa"],"score":75.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1350},"family":"Terminal-Bench-Science","row_ids":["5555c117d96b482e6d1bb0b304cbbefd"],"score":71.05},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":75},"family":"HLE","row_ids":["52db35ab0a7c574170e312e2a679c471"],"score":37.5}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v2","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","peer_count":39,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2}],"score":60.17,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":6,"exact":{"denominator":756,"numerator":37175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":500},"family":"APEX-Agents","row_ids":["100759b239352aa8328473780a99c2b5"],"score":71.43},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":400},"family":"GDPval-AA","row_ids":["8aeb7f071b6b514a7531446104ba33a1"],"score":44.44},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":200},"family":"Terminal-Bench","row_ids":["340dbce590b6aaaa8e1dc907ce4c05ef"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":125},"family":"AutomationBench","row_ids":["29e83d6eda99c0b01ee6e4a2561a6131"],"score":62.5},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":250},"family":"AA-Briefcase","row_ids":["02e903393ff93c77cfd706e931424b8b"],"score":27.78},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":200},"family":"MCP Atlas","row_ids":["938965e9ee08fb381a111173123203b2"],"score":22.22}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","included":true,"model_id":"m-0510680735d3c79a","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"2.1","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"4.0","peer_count":49,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v1.1","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8}],"score":49.17,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["ff5c1b3b8cf739706c1637caa41bfd3b"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1.1","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":1,"numerator":70},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"IFBench","row_ids":["3cf0fe91d3392b3852bb737856d256d6"],"score":40.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"GDP.pdf","row_ids":["5f7d8e688eab4e132d7071fdcec04c3f"],"score":100.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7}],"score":70.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":81,"numerator":5600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1100},"family":"Humanity Sixth Sense","row_ids":["8a450f924e67dc0b617d0ac58b1154ec"],"score":40.74},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":100},"family":"MMMU-Pro","row_ids":["267266c10eb3ebd28421d19f9b4fb8da"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"Chartography","row_ids":["c10b5e0559e0dd2ee1b376a9b0908819"],"score":66.67}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","peer_count":28,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9}],"score":69.14,"weight":10}],"eligible":true,"id":"m-0510680735d3c79a","in_window":true,"name":"GPT-5.6 Sol","official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profile_id":"native","rank":26,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"2e3935359cea7e94134e77b2b6c16a83","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v2","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":83.0,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":83.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"reasoning","direction":"higher_better","id":"06022591029d9808b7b2c118e22277aa","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Sol. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":94.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"94.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":94.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5555c117d96b482e6d1bb0b304cbbefd","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"0.1.0","peer_count":39,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":20.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"20.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row12"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row12","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":20.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"gpt-5.6-sol","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"52db35ab0a7c574170e312e2a679c471","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Diamond","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":31.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7f031bcb1a95e5dbd6cf5720"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"7f031bcb1a95e5dbd6cf5720","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":31.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"100759b239352aa8328473780a99c2b5","included":true,"model_id":"m-0510680735d3c79a","peer_count":8,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":37.9,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"37.9","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / GPT-5.6 Sol"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.9},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T03:27:19.907965+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"agents","direction":"higher_better","feed_id":"public-release-0dd3bed5c7dd1c27","id":"8aeb7f071b6b514a7531446104ba33a1","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"2.1","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Published page official-0dd3bed5c7dd; row Knowledge work GDPval-AA v2.1; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1588.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"fb55426dc69ad840b8b3945c6bbc3ae4b8f5f7b5a2cd69cecf17811da8cc4770","response_sha256":"35b04aa5c4e4556f66c471d48d30adfff181563c76aab2fe07830e54d615831e","result_date":"2026-09-22","source_cell":"1588","source_id":"s-b7fe1fad59fee5842bed","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-0dd3bed5c7dd:table0:row8:column5"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-0dd3bed5c7dd:table0:row8:column5","source_url":"https://www.anthropic.com/claude-opus-5-5","unit":"分","value":1588.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Sol / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"340dbce590b6aaaa8e1dc907ce4c05ef","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"4.0","peer_count":49,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.88,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.88%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row17","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.88},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"agents","direction":"higher_better","id":"29e83d6eda99c0b01ee6e4a2561a6131","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published developer protocol","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":18.1,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table1:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"gpt6-astra-official:table1:row1:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":18.1},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Sol (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"02e903393ff93c77cfd706e931424b8b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"v1.1","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3181818181818182,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6729625931122b9a371676c"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"d6729625931122b9a371676c","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3181818181818182},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"gpt-5.6 (sol)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"938965e9ee08fb381a111173123203b2","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":81.8,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f3a95cd6d81885c3cb8a07bc"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"f3a95cd6d81885c3cb8a07bc","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":81.8},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"coding","direction":"higher_better","id":"ff5c1b3b8cf739706c1637caa41bfd3b","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"1.1","peer_count":9,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":72.7,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column2","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":72.7},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"GPT5.6 Sol (max)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3cf0fe91d3392b3852bb737856d256d6","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":72.7,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"72.7","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column3"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column3","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":72.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"language_information","direction":"higher_better","id":"5f7d8e688eab4e132d7071fdcec04c3f","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current published protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":30.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":30.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Sol (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"8a450f924e67dc0b617d0ac58b1154ec","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"current public protocol","peer_count":28,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":30,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"082c358555d456ed69ad9882"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"082c358555d456ed69ad9882","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":30},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","id":"267266c10eb3ebd28421d19f9b4fb8da","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"Standard 10","peer_count":3,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":84.6,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column1"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column1","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"GPT-5.6 Sol","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"c10b5e0559e0dd2ee1b376a9b0908819","included":true,"model_id":"m-0510680735d3c79a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":79.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"79.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column2","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":79.9}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":60.75,"score_exact":{"denominator":344736,"numerator":20942839},"source_id":"openai/gpt-5.6-sol"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":26334,"numerator":1972813},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":700},"family":"GDPval-AA","row_ids":["efd709c75b89ba20347f69709bb9a3a4"],"score":77.78},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":78},"family":"Terminal-Bench","row_ids":["070c8b1a9e5fb50b3ddeb6aba50faf78"],"score":78.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1500},"family":"AutomationBench","row_ids":["712bffe9ecc8daeca7e0dc7a1fe07e58"],"score":78.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5000},"family":"AA-Briefcase","row_ids":["d6fbe929d324cc028c3c012f923590ba"],"score":64.94}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","peer_count":82,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","peer_count":76,"percentile":78.0,"percentile_exact":{"denominator":1,"numerator":78},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","peer_count":78,"percentile":64.94,"percentile_exact":{"denominator":77,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94}],"score":74.92,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":38,"numerator":1625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1625},"family":"GDP.pdf","row_ids":["d7e50c008783e8a4b24ebddf0211490f"],"score":42.76}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"score":42.76,"weight":20}],"eligible":true,"id":"m-f74476b1532b3a4f","in_window":true,"name":"Ling 3.1 Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":27,"records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"efd709c75b89ba20347f69709bb9a3a4","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v2.1","peer_count":82,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1626.38,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-1-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1626.38},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"070c8b1a9e5fb50b3ddeb6aba50faf78","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"4.0","peer_count":76,"percentile":78.0,"percentile_exact":{"denominator":1,"numerator":78},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.333333333333333,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-1-flash:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.333333333333333},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"712bffe9ecc8daeca7e0dc7a1fe07e58","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":78.95,"percentile_exact":{"denominator":19,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6174373352506193,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-1-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6174373352506193},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d6fbe929d324cc028c3c012f923590ba","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"v1.1","peer_count":78,"percentile":64.94,"percentile_exact":{"denominator":77,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1396.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-1-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1396.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.1 Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e50c008783e8a4b24ebddf0211490f","included":true,"model_id":"m-f74476b1532b3a4f","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-1-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-1-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"release_date":"2026-10-01","release_status":"published","release_verified":true,"result_updated_at":null,"score":60.63,"score_exact":{"denominator":237006,"numerator":14368565},"source_id":"inclusionai/ling-3.1-flash"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":1200591,"numerator":76127425},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5100},"family":"GPQA Diamond","row_ids":["85f2af62bad122fb48d79f2a5cd0ba76"],"score":86.44},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["6cbde617319ab4da7917bf56251c3a2e"],"score":44.74},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3700},"family":"HLE","row_ids":["20084307dcc6cac0bc9c76758b5e2139"],"score":58.73},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3250},"family":"CritPt","row_ids":["6177f516623a8dfc4d1c35316f9c20f2"],"score":63.73}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":86.44,"percentile_exact":{"denominator":59,"numerator":5100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","peer_count":64,"percentile":58.73,"percentile_exact":{"denominator":63,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335}],"score":63.41,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":4,"exact":{"denominator":196240968,"numerator":14270932625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":250},"family":"APEX-Agents","row_ids":["21a8aa59fcb39de6e961594401b8acf8"],"score":83.33},{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":100},"family":"tau3-Banking","row_ids":["0034bad285685afff6ae0af098ac3976"],"score":100.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":7000},"family":"GDPval-AA","row_ids":["8faa2a90e5a7df630665e90e7dc8da8c"],"score":86.42},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1675},"family":"Terminal-Bench","row_ids":["379465a1dce0780ec05eea7fd360c800"],"score":69.79},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1525},"family":"AutomationBench","row_ids":["a4b22debd5eda444528d12c717ed41a1"],"score":80.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6000},"family":"AA-Briefcase","row_ids":["98953845b6332ba7948a61f814a750d3"],"score":77.92},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["4a39f12346421fa0779ea6e76813ff8f"],"score":33.33},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":700},"family":"MCP Atlas","row_ids":["442d36ffd8ae4e2ebc285d09f51efdf2"],"score":77.78},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1050},"family":"AA-AnalystAgent","row_ids":["18b8a0177c3660f9aaf285f2b3571603"],"score":45.65}],"family_count":9,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","peer_count":82,"percentile":86.42,"percentile_exact":{"denominator":81,"numerator":7000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":78,"percentile":77.92,"percentile_exact":{"denominator":77,"numerator":6000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875}],"score":72.72,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":273,"numerator":16000},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":400},"family":"FrontierCode","row_ids":["326d68113ecd602bd7d85d3d2b454d3f"],"score":57.14},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":800},"family":"Vals Code Migration","row_ids":["78dbe912ae9dbb4dfacad05113fef49e"],"score":61.54},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":400},"family":"DeepSWE","row_ids":["a1f523e170da53cbcf5ad3cbcb817e6a"],"score":57.14}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"2.0","peer_count":50,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1.1","peer_count":27,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","included":true,"model_id":"m-c65e3e17492500e9","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1}],"score":58.61,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":34238,"numerator":1354975},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["ba325e0068d64ba19ce176fe95dbb3e4"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":400},"family":"SimpleQA Verified","row_ids":["dd5c55d1f80c2ba62c676b7e3e150a64"],"score":23.53},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2500},"family":"AA-Omniscience","row_ids":["be51199fdc7d1731045d427201852660","330a13ca3aada5b9674ba75668e0acfc"],"score":47.17}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838}],"score":39.58,"weight":20}],"eligible":true,"id":"m-c65e3e17492500e9","in_window":true,"name":"GLM-5.3","official_evidence":{"artifact_sha256":"2dd162815ce7015e44d62c786bf3b76ffde8b7dc7834b1cd6a791f78c3d68667","basis":"official_release_page","date":"2026-08-14","sha256":"2dd162815ce7015e44d62c786bf3b76ffde8b7dc7834b1cd6a791f78c3d68667","url":"https://www.zhipuai.cn/zh/research/162"},"primary_profile_id":"native","profile_id":"native","rank":28,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85f2af62bad122fb48d79f2a5cd0ba76","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":86.44,"percentile_exact":{"denominator":59,"numerator":5100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.917171717171717,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.917171717171717},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6cbde617319ab4da7917bf56251c3a2e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20084307dcc6cac0bc9c76758b5e2139","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Diamond","peer_count":64,"percentile":58.73,"percentile_exact":{"denominator":63,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.179,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6177f516623a8dfc4d1c35316f9c20f2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2 Hard","peer_count":52,"percentile":63.73,"percentile_exact":{"denominator":51,"numerator":3250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.335,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.335},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"GLM-5.3 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"21a8aa59fcb39de6e961594401b8acf8","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"Current published protocol","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":38.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:GLM-5.3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:GLM-5.3","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":38.1},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0034bad285685afff6ae0af098ac3976","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.503092783505155,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.503092783505155},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8faa2a90e5a7df630665e90e7dc8da8c","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v2.1","peer_count":82,"percentile":86.42,"percentile_exact":{"denominator":81,"numerator":7000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1656.88,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1656.88},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"379465a1dce0780ec05eea7fd360c800","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"4.0","peer_count":49,"percentile":69.79,"percentile_exact":{"denominator":24,"numerator":1675},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":38.89,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"38.89%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":38.89},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a4b22debd5eda444528d12c717ed41a1","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":80.26,"percentile_exact":{"denominator":19,"numerator":1525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.622028649962642,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.622028649962642},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"98953845b6332ba7948a61f814a750d3","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":78,"percentile":77.92,"percentile_exact":{"denominator":77,"numerator":6000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1502.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1502.84},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a39f12346421fa0779ea6e76813ff8f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA run","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.36436884512085943,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436884512085943},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"GLM 5.3","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"442d36ffd8ae4e2ebc285d09f51efdf2","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":84.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c780cb0a58d69ac4b6db821f"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"c780cb0a58d69ac4b6db821f","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":84.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"18b8a0177c3660f9aaf285f2b3571603","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA pass^5","peer_count":24,"percentile":45.65,"percentile_exact":{"denominator":23,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.3875,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3875},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"326d68113ecd602bd7d85d3d2b454d3f","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"2.0","peer_count":50,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.21587095959596,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"glm-5-3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21587095959596},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"78dbe912ae9dbb4dfacad05113fef49e","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"1.1","peer_count":27,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":55.07,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"55.07%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row11"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row11","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":55.07},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · GLM-5.3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"a1f523e170da53cbcf5ad3cbcb817e6a","included":true,"model_id":"m-c65e3e17492500e9","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":68.1,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"68.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / GLM-5.3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":68.1},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ba325e0068d64ba19ce176fe95dbb3e4","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dd5c55d1f80c2ba62c676b7e3e150a64","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"run 1.2.0","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.41,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-28","result_timestamp":"2026-08-28T17:00:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ad9049470ee6f9a7ee342d1d"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"ad9049470ee6f9a7ee342d1d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.41},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"be51199fdc7d1731045d427201852660","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":50.94,"percentile_exact":{"denominator":53,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"330a13ca3aada5b9674ba75668e0acfc","included":true,"model_id":"m-c65e3e17492500e9","original_version_label":"v1.1","peer_count":54,"percentile":43.4,"percentile_exact":{"denominator":53,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6113300492610838,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6113300492610838}],"release_date":"2026-08-14","release_status":"published","release_verified":true,"result_updated_at":"2026-08-28","score":59.63,"score_exact":{"denominator":2441081826591408,"numerator":145567376505820175},"source_id":"zhipuai/glm-5.3"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":40698,"numerator":2570275},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":350},"family":"FrontierMath","row_ids":["21e96013106dd82d62904afe0c4f768b"],"score":38.89},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["0e248a56228a0b9ac7b762f2f3b00fc9"],"score":57.89},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":5000},"family":"HLE","row_ids":["de9339207cccf727d5f73de793318c18"],"score":79.37},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":1300},"family":"CritPt","row_ids":["b0085122202e14e07e4a784b0954985f"],"score":76.47}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Diamond","peer_count":64,"percentile":79.37,"percentile_exact":{"denominator":63,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","peer_count":52,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435}],"score":63.15,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":143451,"numerator":9916043},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":2200},"family":"GDPval-AA","row_ids":["ac3f66c4919122d1f31985187718ed07"],"score":81.48},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":224},"family":"Terminal-Bench","row_ids":["7706b3202bdd1e09b6c5e4786d4bb3e4"],"score":74.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"AutomationBench","row_ids":["9f3e53e10beed094c8835d4e672cc44f"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6800},"family":"AA-Briefcase","row_ids":["6fc39f46271fccbf5e9ec9022c726f0f"],"score":88.31},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":100},"family":"OSWorld","row_ids":["5b785287811a33de601a7060b80bab3e"],"score":33.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":2000},"family":"AA-AnalystAgent","row_ids":["a7b164121b0a9a4769bf158d9212d770"],"score":86.96}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2.1","peer_count":82,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":78,"percentile":88.31,"percentile_exact":{"denominator":77,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.1","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","peer_count":24,"percentile":86.96,"percentile_exact":{"denominator":23,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375}],"score":69.12,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":91,"numerator":6350},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":600},"family":"FrontierCode","row_ids":["cbf1fa28fdc821e9836816e2dbc59c73"],"score":85.71},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":700},"family":"Vals Code Migration","row_ids":["c1b5961685ee27b7ccd1b99becf1d743"],"score":53.85}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.0","peer_count":50,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"1.1","peer_count":27,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36}],"score":69.78,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":4819875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2975},"family":"GDP.pdf","row_ids":["a183b88682c41fc78cc2ad5e5485bbb7"],"score":78.29},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":100},"family":"SimpleQA Verified","row_ids":["a7e4020697adbd00e97487d7c55696e0"],"score":5.88},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"MultiChallenge","row_ids":["5d9b045cf9ce01ea061077e1b2e84c1e"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3000},"family":"AA-Omniscience","row_ids":["30bc019f7dc3e7314870de193c23616d","b11d60a588bf725879871b88bc8f70cb"],"score":56.6}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897}],"score":35.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"Chartography","row_ids":["888dd585273ea6255f0bebeb11182076"],"score":50.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4}],"score":50.0,"weight":10}],"eligible":true,"id":"m-016da3c7828e329e","in_window":true,"name":"Claude Haiku 5.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-07","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":29,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"21e96013106dd82d62904afe0c4f768b","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.4634146341463415,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-07","result_timestamp":"2026-10-07T20:42:31.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8b0aae3ea9b398a4d6507285"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"8b0aae3ea9b398a4d6507285","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4634146341463415},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0e248a56228a0b9ac7b762f2f3b00fc9","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"de9339207cccf727d5f73de793318c18","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"Diamond","peer_count":64,"percentile":79.37,"percentile_exact":{"denominator":63,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.296,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-haiku-5-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.296},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b0085122202e14e07e4a784b0954985f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2 Hard","peer_count":52,"percentile":76.47,"percentile_exact":{"denominator":17,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.435,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-haiku-5-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.435},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ac3f66c4919122d1f31985187718ed07","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v2.1","peer_count":82,"percentile":81.48,"percentile_exact":{"denominator":27,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1630.83,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-haiku-5-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1630.83},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7706b3202bdd1e09b6c5e4786d4bb3e4","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"4.0","peer_count":76,"percentile":74.67,"percentile_exact":{"denominator":3,"numerator":224},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.328282828282828,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-haiku-5-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.328282828282828},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9f3e53e10beed094c8835d4e672cc44f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.35411015132783624,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-haiku-5-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35411015132783624},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6fc39f46271fccbf5e9ec9022c726f0f","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":78,"percentile":88.31,"percentile_exact":{"denominator":77,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1574.12,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-haiku-5-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1574.12},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"5b785287811a33de601a7060b80bab3e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.1","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":72.4,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"72.4% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":72.4},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7b164121b0a9a4769bf158d9212d770","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA pass^5","peer_count":24,"percentile":86.96,"percentile_exact":{"denominator":23,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.5375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"claude-haiku-5-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"cbf1fa28fdc821e9836816e2dbc59c73","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"2.0","peer_count":50,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.349325505050505,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-haiku-5-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.349325505050505},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Claude Haiku 5.5 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c1b5961685ee27b7ccd1b99becf1d743","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"1.1","peer_count":27,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":52.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"52.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row13","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":52.36},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a183b88682c41fc78cc2ad5e5485bbb7","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current AA protocol","peer_count":77,"percentile":78.29,"percentile_exact":{"denominator":38,"numerator":2975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.208,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-haiku-5-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-haiku-5-5_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"a7e4020697adbd00e97487d7c55696e0","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"run 1.2.0","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.238,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:36:06.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d3909a3e4cc15e1a51a0db6"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"4d3909a3e4cc15e1a51a0db6","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.238},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Claude Haiku 5.5","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"5d9b045cf9ce01ea061077e1b2e84c1e","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"current LLM protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":67.67,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4d1f0c8fd5c56c69c34068d7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"4d1f0c8fd5c56c69c34068d7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":67.67},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30bc019f7dc3e7314870de193c23616d","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":54.72,"percentile_exact":{"denominator":53,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3466666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3466666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Claude Haiku 5.5 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b11d60a588bf725879871b88bc8f70cb","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6576530612244897,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-haiku-5-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-haiku-5-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6576530612244897},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"Haiku 5.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"888dd585273ea6255f0bebeb11182076","included":true,"model_id":"m-016da3c7828e329e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":46.4,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"46.4% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column1"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column1","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":46.4}],"release_date":"2026-10-07","release_status":"published","release_verified":true,"result_updated_at":"2026-10-09","score":59.06,"score_exact":{"denominator":127698358788,"numerator":7542473801321},"source_id":"anthropic/claude-haiku-5-5"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":1,"numerator":50},"family":"Terminal-Bench-Science","row_ids":["174d2abbe0e35fdfac2a047ae2fa2208"],"score":50.0}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","peer_count":39,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14}],"score":50.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":72,"numerator":2375},"families":[{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":325},"family":"Terminal-Bench","row_ids":["d7c99e26bef72e8691defabf85dff136"],"score":27.08},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":350},"family":"AA-Briefcase","row_ids":["d43d801b58cb460394b97c15079fc130"],"score":38.89}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","peer_count":49,"percentile":27.08,"percentile_exact":{"denominator":12,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365}],"score":32.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"run 0.1.200","exact":{"denominator":2,"numerator":125},"family":"MirrorCode","row_ids":["02fcbee985f3da4b53d62ce022f2a290"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1400},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1400},"family":"SimpleQA Verified","row_ids":["ab75c5704ac20fa5dc1c93828558d2ce"],"score":82.35}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692}],"score":82.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":27,"numerator":2000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":2000},"family":"Humanity Sixth Sense","row_ids":["e0d9bd2a6ca55a94058a8fdfafb19d29"],"score":74.07}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","peer_count":28,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8}],"score":74.07,"weight":10}],"eligible":true,"id":"m-16a8dce2a6e7167a","in_window":true,"name":"Gemini 3.7 Flash","official_evidence":{"artifact_sha256":"b68cd4829f86d5c5a5cafdbccf2e60b6956a5bfc04074d538c1dd7d5c4a9a559","basis":"official_release_page","date":"2026-08-13","sha256":"b68cd4829f86d5c5a5cafdbccf2e60b6956a5bfc04074d538c1dd7d5c4a9a559","url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash/"},"primary_profile_id":"native","profile_id":"native","rank":30,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"174d2abbe0e35fdfac2a047ae2fa2208","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"0.1.0","peer_count":39,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":7.14,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"7.14%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row21"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row21","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":7.14},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.7 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d7c99e26bef72e8691defabf85dff136","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"4.0","peer_count":49,"percentile":27.08,"percentile_exact":{"denominator":12,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.12,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"12.12%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":12.12},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.7 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"d43d801b58cb460394b97c15079fc130","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"v1.1","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.36363636363636365,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aeba34aaf766d3d8633eddbe"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"aeba34aaf766d3d8633eddbe","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.36363636363636365},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"02fcbee985f3da4b53d62ce022f2a290","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:08.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4c7a7df2018d27f949be46f"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f4c7a7df2018d27f949be46f","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.7-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"ab75c5704ac20fa5dc1c93828558d2ce","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"run 1.2.0","peer_count":35,"percentile":82.35,"percentile_exact":{"denominator":17,"numerator":1400},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.692,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"aae21e83c2f475930bafb6c9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"aae21e83c2f475930bafb6c9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.692},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.7-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"e0d9bd2a6ca55a94058a8fdfafb19d29","included":true,"model_id":"m-16a8dce2a6e7167a","original_version_label":"current public protocol","peer_count":28,"percentile":74.07,"percentile_exact":{"denominator":27,"numerator":2000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":39.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f4dfd0da2336f564905e98a8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f4dfd0da2336f564905e98a8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":39.8}],"release_date":"2026-08-13","release_status":"published","release_verified":true,"result_updated_at":"2026-10-21","score":57.12,"score_exact":{"denominator":14688,"numerator":839045},"source_id":"google/gemini-3.7-flash"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":4200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4200},"family":"GPQA Diamond","row_ids":["004e3b6a13ca6828e88e6f4e76a97e76"],"score":71.19}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":71.19,"percentile_exact":{"denominator":59,"numerator":4200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192}],"score":71.19,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3081078,"numerator":117236125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["0456438c8b02eed2e908e4593e41dcf0"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3700},"family":"GDPval-AA","row_ids":["9050cb18d0244c05aa4a3bdecef069eb"],"score":45.68},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":700},"family":"AutomationBench","row_ids":["c9e2b1db45e7b21cbd9a431702d99eb1"],"score":36.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2700},"family":"AA-Briefcase","row_ids":["82a245d0d5111b9661f660b2a94b3db6"],"score":35.06}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v2.1","peer_count":82,"percentile":45.68,"percentile_exact":{"denominator":81,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v1.1","peer_count":78,"percentile":35.06,"percentile_exact":{"denominator":77,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26}],"score":38.05,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":125},"family":"IFBench","row_ids":["353142d52584e4a6aac9d52541e06d23"],"score":62.5},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":2225},"family":"GDP.pdf","row_ids":["312a7fa68250f3fcf7a32404595e6dc4"],"score":58.55}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126}],"score":60.53,"weight":20}],"eligible":true,"id":"m-c0755395b2da9721","in_window":true,"name":"Nex-N2-Pro","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-02","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":31,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"004e3b6a13ca6828e88e6f4e76a97e76","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":71.19,"percentile_exact":{"denominator":59,"numerator":4200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.891919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nex-n2-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.891919191919192},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0456438c8b02eed2e908e4593e41dcf0","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nex-n2-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9050cb18d0244c05aa4a3bdecef069eb","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v2.1","peer_count":82,"percentile":45.68,"percentile_exact":{"denominator":81,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.48,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nex-n2-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.48},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9e2b1db45e7b21cbd9a431702d99eb1","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17629116737896317,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nex-n2-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17629116737896317},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82a245d0d5111b9661f660b2a94b3db6","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"v1.1","peer_count":78,"percentile":35.06,"percentile_exact":{"denominator":77,"numerator":2700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":940.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nex-n2-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":940.26},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"353142d52584e4a6aac9d52541e06d23","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current 294-item protocol","peer_count":17,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.661904761904762,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nex-n2-pro:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.661904761904762},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nex-N2-Pro (Based on Qwen3.5-397B-A17B)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"312a7fa68250f3fcf7a32404595e6dc4","included":true,"model_id":"m-c0755395b2da9721","original_version_label":"current AA protocol","peer_count":77,"percentile":58.55,"percentile_exact":{"denominator":38,"numerator":2225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.126,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nex-n2-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nex-n2-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126}],"release_date":"2026-06-02","release_status":"published","release_verified":true,"result_updated_at":null,"score":56.31,"score_exact":{"denominator":2544970428,"numerator":143298061675},"source_id":"nex-agi/nex-n2-pro"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":7075},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":175},"family":"GPQA Diamond","row_ids":["7ff900665f3a2edd6bd47eeb54dcc81e"],"score":58.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1250},"family":"Terminal-Bench-Science","row_ids":["a18d9b692129832a6701c8778a0404f3"],"score":65.79}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":58.33,"percentile_exact":{"denominator":3,"numerator":175},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","peer_count":39,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86}],"score":62.06,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":1008,"numerator":58825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":200},"family":"APEX-Agents","row_ids":["bf4bf643f4907b71184823fd0ebd6b1c"],"score":28.57},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":725},"family":"Terminal-Bench","row_ids":["e085a37b9093e9f4b5222ae848bc472f"],"score":60.42},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"AutomationBench","row_ids":["58a53dd6f007e225a58569d30c64eee8"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":850},"family":"AA-Briefcase","row_ids":["0d1d30147df15075a1ed15467a0ef1b4"],"score":94.44}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","included":true,"model_id":"m-99ca714b8835cfb9","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","peer_count":49,"percentile":60.42,"percentile_exact":{"denominator":12,"numerator":725},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"reference_revision":"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593}],"score":58.36,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["9619717553c9af1d9fdaa3ac01eac67f"],"score":25.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"reference_revision":"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6}],"score":25.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":17,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["57da33d0598ebe148085f216c7c7cefb"],"score":100.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":600},"family":"SimpleQA Verified","row_ids":["2e690ecc6ebeb390f087e351bd8c7803"],"score":35.29}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458}],"score":67.65,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":54,"numerator":3725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1700},"family":"Humanity Sixth Sense","row_ids":["b0de0689b04f1b90ca42e94e49e0a3bb"],"score":62.96},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":75},"family":"MMMU-Pro","row_ids":["d1f2e18c925df69cd229259da568a231"],"score":75.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","peer_count":28,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3}],"score":68.98,"weight":10}],"eligible":true,"id":"m-99ca714b8835cfb9","in_window":true,"name":"Qwen3.8 Max","official_evidence":{"artifact_sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","basis":"official_release_page","date":"2026-08-02","sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","url":"https://www.alibabacloud.com/help/en/model-studio/newly-released-models"},"primary_profile_id":"native","profile_id":"native","rank":32,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"7ff900665f3a2edd6bd47eeb54dcc81e","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":58.33,"percentile_exact":{"denominator":3,"numerator":175},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.8-Max. Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":92.6,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"92.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row25:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-weights:table0:row25:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"a18d9b692129832a6701c8778a0404f3","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"0.1.0","peer_count":39,"percentile":65.79,"percentile_exact":{"denominator":19,"numerator":1250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":12.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"12.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":12.86},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"bf4bf643f4907b71184823fd0ebd6b1c","included":true,"model_id":"m-99ca714b8835cfb9","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":34.0,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"34.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Qwen3.8 Max"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":34.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Qwen3.8 Max / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"e085a37b9093e9f4b5222ae848bc472f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"4.0","peer_count":49,"percentile":60.42,"percentile_exact":{"denominator":12,"numerator":725},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":34.34,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"34.34%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row20","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":34.34},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"58a53dd6f007e225a58569d30c64eee8","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"published developer protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-weights; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"2a48ec513964ac2d123d226b","raw_value":27.3,"reference_revision":"6e951399a962e87cb2fbdd7aeb3931277f62e590eefb13d0527dcf7fed6a4541","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"27.3","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row20:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-weights:table0:row20:column5","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":27.3},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.8 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0d1d30147df15075a1ed15467a0ef1b4","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"v1.1","peer_count":19,"percentile":94.44,"percentile_exact":{"denominator":9,"numerator":850},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.49595959595959593,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf43f1e7067ff45325926c2b"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"cf43f1e7067ff45325926c2b","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.49595959595959593},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"9619717553c9af1d9fdaa3ac01eac67f","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page qwen38-max; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"eb23265b5e5c19deb4dff9c1","raw_value":56.6,"reference_revision":"b419a7444f4958e41bfc7f2db965fe23d5b3a0d5c6b64f991fda2bdbe6299bad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"56.6","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row4:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-max:table1:row4:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":56.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"57da33d0598ebe148085f216c7c7cefb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":82.8,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.8","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column5"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column5","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.8},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.8-max_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2e690ecc6ebeb390f087e351bd8c7803","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"run 1.2.0","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.458,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:55.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bc3e088af396775fcc497753"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"bc3e088af396775fcc497753","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.458},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Qwen3.8-Max (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b0de0689b04f1b90ca42e94e49e0a3bb","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"current public protocol","peer_count":28,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":33,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8de26ec1a7320e0e6d80daec"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8de26ec1a7320e0e6d80daec","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":33},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.8-Max","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"d1f2e18c925df69cd229259da568a231","included":true,"model_id":"m-99ca714b8835cfb9","original_version_label":"Standard 10","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":82.3,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"82.3","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column6","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":82.3}],"release_date":"2026-08-02","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":55.53,"score_exact":{"denominator":3907008,"numerator":216965705},"source_id":"alibaba/qwen3.8-max"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":19,"numerator":1200},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1200},"family":"Terminal-Bench-Science","row_ids":["545c704b3f9b5b5c3a00fbe6211a666c"],"score":63.16}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","peer_count":39,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43}],"score":63.16,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":144,"numerator":6875},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":275},"family":"Terminal-Bench","row_ids":["620e1ee85f651d96e2f6e21d676ecc23"],"score":34.38},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":550},"family":"AA-Briefcase","row_ids":["45f7b80c7cc89c267cb8e0d74593b6e4"],"score":61.11}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202}],"score":47.74,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"run 0.1.200","exact":{"denominator":2,"numerator":125},"family":"MirrorCode","row_ids":["0621e61b8874eac3232c0422d1f71677"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111}],"score":62.5,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":17,"numerator":800},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":800},"family":"SimpleQA Verified","row_ids":["954b3bb3d889d2d8fa0f84b9a3223eaf"],"score":47.06}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493}],"score":47.06,"weight":20}],"eligible":true,"id":"m-3c17c0d3ca7f1abe","in_window":true,"name":"Grok 4.6","official_evidence":{"artifact_sha256":"22a9c3dcee05d0c5a93b52c78e1c1e867278c4f39ebda057faec281332c01411","basis":"official_release_page","date":"2026-08-12","sha256":"22a9c3dcee05d0c5a93b52c78e1c1e867278c4f39ebda057faec281332c01411","url":"https://x.ai/news/grok-4-6"},"primary_profile_id":"native","profile_id":"native","rank":33,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"545c704b3f9b5b5c3a00fbe6211a666c","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"0.1.0","peer_count":39,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"11.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row15","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":11.43},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.6 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"620e1ee85f651d96e2f6e21d676ecc23","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row35"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row35","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.6 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"45f7b80c7cc89c267cb8e0d74593b6e4","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"v1.1","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.402020202020202,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0fe184e0482b5b4211f30726"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"0fe184e0482b5b4211f30726","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.402020202020202},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_xhigh","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0621e61b8874eac3232c0422d1f71677","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 0.1.200","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.4111111111111111,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:25.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f30e81f1c591345614707515"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"f30e81f1c591345614707515","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.4111111111111111},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.6_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"954b3bb3d889d2d8fa0f84b9a3223eaf","included":true,"model_id":"m-3c17c0d3ca7f1abe","original_version_label":"run 1.2.0","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.493,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:53.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"dccbd5b9b2f77df874e6d9f9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"dccbd5b9b2f77df874e6d9f9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.493}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":"2026-10-21","score":55.15,"score_exact":{"denominator":837216,"numerator":46174325},"source_id":"xai/grok-4.6"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1642914,"numerator":100299025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4900},"family":"GPQA Diamond","row_ids":["3551c30b97ba4cb5f0fc94e087d94b59"],"score":83.05},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":750},"family":"Terminal-Bench-Science","row_ids":["0ca35771477057319be8d71ddeaad930"],"score":57.69},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3800},"family":"HLE","row_ids":["a57cb2704c93293d339e91362f40ae46"],"score":60.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2200},"family":"CritPt","row_ids":["620b6f3e6314727d16068f5503869955"],"score":43.14}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":83.05,"percentile_exact":{"denominator":59,"numerator":4900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","peer_count":64,"percentile":60.32,"percentile_exact":{"denominator":63,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","peer_count":52,"percentile":43.14,"percentile_exact":{"denominator":51,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245}],"score":61.05,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":141729588,"numerator":8766837125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":2375},"family":"tau3-Banking","row_ids":["9a44264bd8b89c341b6a8bd9496be3bf"],"score":91.35},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6800},"family":"GDPval-AA","row_ids":["3091d538e8d1e5edf6eea60314cea6a6"],"score":83.95},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":775},"family":"Terminal-Bench","row_ids":["28abab0f2a19e622cf15ffcdc9be3dfd"],"score":64.58},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1475},"family":"AutomationBench","row_ids":["e492bf3e10afd1293a1152cd005f96a3"],"score":77.63},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":800},"family":"AA-Briefcase","row_ids":["3015c7ad3cd2d55a621f41dcebd98f3d"],"score":72.73},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["b8b704019523524af0ce859e89927fa0"],"score":16.67},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":600},"family":"AA-AnalystAgent","row_ids":["a9057fe2d742aeb1aa0666773aa5958d"],"score":26.09}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","peer_count":82,"percentile":83.95,"percentile_exact":{"denominator":81,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","peer_count":49,"percentile":64.58,"percentile_exact":{"denominator":12,"numerator":775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":77.63,"percentile_exact":{"denominator":19,"numerator":1475},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":78,"percentile":72.73,"percentile_exact":{"denominator":11,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","peer_count":24,"percentile":26.09,"percentile_exact":{"denominator":23,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25}],"score":61.86,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":30950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2500},"family":"FrontierCode","row_ids":["32536ff0b3b8576673f9f05323026239"],"score":51.02},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":600},"family":"Vals Code Migration","row_ids":["efd99741d3524e2d5ab3b679f8da8006"],"score":46.15}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","peer_count":50,"percentile":51.02,"percentile_exact":{"denominator":49,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","peer_count":27,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65}],"score":48.59,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":108625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1225},"family":"GDP.pdf","row_ids":["f95dbe30b012b632cab81f4598931983"],"score":64.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["a76a9fe9ce50e17717765fcaebd6fbf8","7e86ee60903a41ed29953982d04673a1"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499}],"score":53.93,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["7d2a87235efadfc965d04473b687dd29"],"score":33.33}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","peer_count":28,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8}],"score":33.33,"weight":10}],"eligible":true,"id":"m-cf177eb625d4bd3e","in_window":true,"name":"GLM-5.3-Flash","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":34,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3551c30b97ba4cb5f0fc94e087d94b59","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":83.05,"percentile_exact":{"denominator":59,"numerator":4900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.912121212121212,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"glm-5-3-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.912121212121212},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0ca35771477057319be8d71ddeaad930","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"0.1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"glm-5-3-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a57cb2704c93293d339e91362f40ae46","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"Diamond","peer_count":64,"percentile":60.32,"percentile_exact":{"denominator":63,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.182,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"glm-5-3-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.182},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"620b6f3e6314727d16068f5503869955","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2 Hard","peer_count":52,"percentile":43.14,"percentile_exact":{"denominator":51,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.245,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"glm-5-3-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.245},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9a44264bd8b89c341b6a8bd9496be3bf","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":91.35,"percentile_exact":{"denominator":26,"numerator":2375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.472164948453608,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"glm-5-3-flash:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.472164948453608},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3091d538e8d1e5edf6eea60314cea6a6","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v2.1","peer_count":82,"percentile":83.95,"percentile_exact":{"denominator":81,"numerator":6800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1649.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"glm-5-3-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1649.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"28abab0f2a19e622cf15ffcdc9be3dfd","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"4.0","peer_count":49,"percentile":64.58,"percentile_exact":{"denominator":12,"numerator":775},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":37.37,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"37.37%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row18","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":37.37},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e492bf3e10afd1293a1152cd005f96a3","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":77.63,"percentile_exact":{"denominator":19,"numerator":1475},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6036862782167782,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"glm-5-3-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6036862782167782},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3015c7ad3cd2d55a621f41dcebd98f3d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":78,"percentile":72.73,"percentile_exact":{"denominator":11,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1450.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"glm-5-3-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1450.94},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b8b704019523524af0ce859e89927fa0","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA run","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3321396598030439,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"glm-5-3-flash:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3321396598030439},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9057fe2d742aeb1aa0666773aa5958d","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA pass^5","peer_count":24,"percentile":26.09,"percentile_exact":{"denominator":23,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.25,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"glm-5-3-flash:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"32536ff0b3b8576673f9f05323026239","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"2.0","peer_count":50,"percentile":51.02,"percentile_exact":{"denominator":49,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.176070454545455,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176070454545455},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GLM-5.3-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"efd99741d3524e2d5ab3b679f8da8006","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"1.1","peer_count":27,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.65,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.65%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row15"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row15","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.65},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f95dbe30b012b632cab81f4598931983","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current AA protocol","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.154,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"glm-5-3-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a76a9fe9ce50e17717765fcaebd6fbf8","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.282,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.282},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GLM-5.3-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e86ee60903a41ed29953982d04673a1","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6165273909006499,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"glm-5-3-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"glm-5-3-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6165273909006499},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GLM-5.3-Flash (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"7d2a87235efadfc965d04473b687dd29","included":true,"model_id":"m-cf177eb625d4bd3e","original_version_label":"current public protocol","peer_count":28,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":26.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cbb86a4b62febc59436c91ee"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"cbb86a4b62febc59436c91ee","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":26.8}],"release_date":"2026-08-26","release_status":"published","release_verified":true,"result_updated_at":null,"score":54.56,"score_exact":{"denominator":210957688717776,"numerator":11510724316142215},"source_id":"zhipuai/glm-5.3-flash"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":399,"numerator":23825},"families":[{"benchmark_version":"2","exact":{"denominator":7,"numerator":600},"family":"FrontierMath","row_ids":["51cb1dea03ece9a899f843aedba7597e"],"score":85.71},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":25},"family":"GPQA Diamond","row_ids":["b6943c0f15a31bf97046c2e4af0b90f3"],"score":25.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1300},"family":"Terminal-Bench-Science","row_ids":["f5cc9d3986b503f55588d97bf5fde40a"],"score":68.42}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","peer_count":8,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","peer_count":39,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71}],"score":59.71,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":144,"numerator":8225},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"APEX-Agents","row_ids":["b82b11faf1fed2434465112eb548316c"],"score":100.0},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":875},"family":"Terminal-Bench","row_ids":["5ecd3796946e727694a868a88027022a"],"score":72.92},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["8ad20dc033b0a09d811fcebfca51c2ab"],"score":0.0},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":500},"family":"MCP Atlas","row_ids":["37801ada9ba8d94f8b7e8f5a6667ee11"],"score":55.56}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","peer_count":49,"percentile":72.92,"percentile_exact":{"denominator":12,"numerator":875},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3}],"score":57.12,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"DeepSWE","row_ids":["0772cd4ebc7a33c01cf749b6a69bd7c0"],"score":50.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","peer_count":9,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9}],"score":50.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":2,"numerator":95},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"IFBench","row_ids":["3f5e26906c244a4d4f9e57725da3d9df"],"score":20.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"GDP.pdf","row_ids":["134a991b936b3bccddf2a9a60c450e69"],"score":75.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8}],"score":47.5,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":3,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["cf924b68bd130c394ca909ce8970cb90"],"score":66.67},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["a6ccc017ec5976d43545f1578c4ba3f8"],"score":50.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","peer_count":28,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2}],"score":58.33,"weight":10}],"eligible":true,"id":"m-86c0c8bf10aa4a0d","in_window":true,"name":"Claude Fable 5","official_evidence":{"artifact_sha256":"c8267145b47f034cb45aedbe8074a7b244fee4ab2050099a2f6e2e7f9d6d9ed5","basis":"official_release_page","date":"2026-06-09","sha256":"c8267145b47f034cb45aedbe8074a7b244fee4ab2050099a2f6e2e7f9d6d9ed5","url":"https://www.anthropic.com/news/claude-fable-5-mythos-5"},"primary_profile_id":"native","profile_id":"native","rank":35,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"51cb1dea03ece9a899f843aedba7597e","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"v2","peer_count":8,"percentile":85.71,"percentile_exact":{"denominator":7,"numerator":600},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":90.2,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"gpt6-astra-official:table3:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":90.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"reasoning","direction":"higher_better","id":"b6943c0f15a31bf97046c2e4af0b90f3","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Fable 5. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.6,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":"92.6%","source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table3:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"gpt6-astra-official:table3:row3:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":92.6},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f5cc9d3986b503f55588d97bf5fde40a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"0.1.0","peer_count":39,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":15.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"15.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row13"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row13","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":15.71},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Fable 5 (max, w/ fallback)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"b82b11faf1fed2434465112eb548316c","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Current published protocol","peer_count":6,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Fable 5 (max, w/ fallback). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":43.3,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"43.3","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column2","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":43.3},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Fable 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5ecd3796946e727694a868a88027022a","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"4.0","peer_count":49,"percentile":72.92,"percentile_exact":{"denominator":12,"numerator":875},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":41.41,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"41.41%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row14","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":41.41},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:35.835921+00:00","component_key":"score","configuration":"Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7232d52f37fd859","id":"8ad20dc033b0a09d811fcebfca51c2ab","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"published developer protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page anthropic-fable51; row Business workflows AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"659066a5e56ba232a06f7c9f","raw_value":17.1,"reference_revision":"1edfa203835b079f61f6108617beb7dda22436621c5eb4783c15eb49794f8e1b","research_artifact_sha256":"ac54a246fe3d374ca612e7c9387e8b10e3162337fe0a98ee10c8d59546c202e7","response_sha256":"3310e6cf7f2ec2efce4b886cdf8a37e3dbdb0243b5fff503e045301d5e4d7a56","result_date":null,"source_cell":"17.1%","source_id":"s-1c3420336981cb6ce894","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"anthropic-fable51:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:AutomationBench:Business workflows AutomationBench","source_record_id":"anthropic-fable51:table0:row15:column2","source_url":"https://www.anthropic.com/claude-fable-and-mythos-5-1","unit":"%","value":17.1},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"37801ada9ba8d94f8b7e8f5a6667ee11","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0a4e67415c213a257bd66bf2"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0a4e67415c213a257bd66bf2","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.3},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"coding","direction":"higher_better","id":"0772cd4ebc7a33c01cf749b6a69bd7c0","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"1.1","peer_count":9,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page gpt6-astra-official; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.9,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"a393c94c187ae0a7bb8027fcc32b810872a845482593d7b546dcfe1913557bd3","response_sha256":"6d2a18325536de3d13adbfea313f75de480d37e191b2d06233ed219d43787571","result_date":"2026-09-29","source_cell":null,"source_id":"s-56f82fed862f1e215c84","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"gpt6-astra-official:table2:row2:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"gpt6-astra-official:table2:row2:column4","source_url":"https://openai.com/index/gpt-6-astra/","unit":"%","value":69.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"3f5e26906c244a4d4f9e57725da3d9df","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":63.5,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"63.5","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":63.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Fable 5","dimension":"language_information","direction":"higher_better","id":"134a991b936b3bccddf2a9a60c450e69","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current published protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":29.8,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column5"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column5","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":29.8},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Fable-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf924b68bd130c394ca909ce8970cb90","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"current public protocol","peer_count":28,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":34.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"93ce9cb4a6c12794d7c977d5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"93ce9cb4a6c12794d7c977d5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":34.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Fable5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"a6ccc017ec5976d43545f1578c4ba3f8","included":true,"model_id":"m-86c0c8bf10aa4a0d","original_version_label":"Standard 10","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page qwen38-max; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"bcb65796e6a9f26048c645d7","raw_value":81.2,"reference_revision":"7d432057105c6c8f9f389404ea12fd175df865828a08aa84f169a69aaf0aeda3","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"81.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table2:row2:column2"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMMU-Pro:MMMU-Pro","source_record_id":"qwen38-max:table2:row2:column2","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":81.2}],"release_date":"2026-06-09","release_status":"published","release_verified":true,"result_updated_at":"2026-09-29","score":54.54,"score_exact":{"denominator":76608,"numerator":4178261},"source_id":"anthropic/claude-fable-5"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["f0716aa67008d67c9923512a4d8c5e11"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":162,"numerator":3275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["01aeec9c08a057635f9b72825df4f3fb"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":50},"family":"tau3-Banking","row_ids":["e24b9f9c121b7281ff7f37b4b404f354"],"score":50.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2500},"family":"GDPval-AA","row_ids":["7bf77397fcea6af7a97a2524f4c194bb"],"score":30.86},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["b96a70ae7b2ded290b5f169ceedf1e03"],"score":0.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","peer_count":82,"percentile":30.86,"percentile_exact":{"denominator":81,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2}],"score":20.22,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":9,"numerator":850},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":800},"family":"IFBench","row_ids":["dafb6731eb5f8725711540078af7013b"],"score":88.89},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"MMLU-Pro","row_ids":["27d4cca7113cd0216eb12449ea9867c2"],"score":100.0}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2}],"score":94.44,"weight":20}],"eligible":true,"id":"m-b7faaf31060e95f7","in_window":true,"name":"Solar Open2 250B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":36,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f0716aa67008d67c9923512a4d8c5e11","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-open2-250b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"01aeec9c08a057635f9b72825df4f3fb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current published protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":16.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"16.6","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"twelve-solar-card:table1:row16:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":16.6},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e24b9f9c121b7281ff7f37b4b404f354","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.216494845360825,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-open2-250b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216494845360825},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Open2 250B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7bf77397fcea6af7a97a2524f4c194bb","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"v2.1","peer_count":82,"percentile":30.86,"percentile_exact":{"denominator":81,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":978,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-open2-250b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-open2-250b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":978},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"agents","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"b96a70ae7b2ded290b5f169ceedf1e03","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":58.2,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"58.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"twelve-solar-card:table1:row17:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":58.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"dafb6731eb5f8725711540078af7013b","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":80.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"80.0","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-solar-card:table1:row11:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":80.0},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:42.022939+00:00","component_key":"score","configuration":"Solar Open 2 250B-A15B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-f16dbde15779b059","id":"27d4cca7113cd0216eb12449ea9867c2","included":true,"model_id":"m-b7faaf31060e95f7","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Solar Open 2 250B-A15B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":86.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"27dc2644ee7bd12124de79f08531d124c11e9530e051f6ac880fe14d125b9147","response_sha256":"b6a8866fa5574f70006e83832da5eccf60d04d47a041d8b8dc746b0fc2e5bb82","result_date":null,"source_cell":"86.2","source_id":"s-0daea7ae978f1cf387ad","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-solar-card:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-solar-card:table1:row2:column1","source_url":"https://huggingface.co/upstage/Solar-Open2-250B","unit":"%","value":86.2}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":54.18,"score_exact":{"denominator":133812,"numerator":7249925},"source_id":"published/aa:solar-open2-250b"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":20349,"numerator":1015100},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":350},"family":"Terminal-Bench-Science","row_ids":["f80ea7a9c8b1ad0f866390a070dec6ae"],"score":18.42},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":500},"family":"HLE","row_ids":["ea28428a53ddb22639d7f5b7e50547a2"],"score":71.43},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3050},"family":"CritPt","row_ids":["b5e83ad92c83f666dbcd3cdc392d63b6"],"score":59.8}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","included":true,"model_id":"m-f892623396612d9b","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","included":true,"model_id":"m-f892623396612d9b","original_version_label":"Diamond","peer_count":64,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325}],"score":49.88,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":152631864,"numerator":9897530725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"APEX-Agents","row_ids":["519c33c19b165c6197391af117e522e3"],"score":75.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5900},"family":"GDPval-AA","row_ids":["d52704f212d466a4d653d53f1794d1db"],"score":72.84},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":1375},"family":"Terminal-Bench","row_ids":["d5a9a67230f3e72676fe2415a004dd47"],"score":57.29},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1175},"family":"AutomationBench","row_ids":["85e95115c24e6ebd5d7a72360c176406"],"score":61.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5300},"family":"AA-Briefcase","row_ids":["5342274720726a814f5f5d822f07b510"],"score":68.83},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":250},"family":"EnterpriseOps-Gym","row_ids":["6be4e7acd5ffa2c31291fe47e51a2a60"],"score":83.33},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":800},"family":"AA-AnalystAgent","row_ids":["1f372ec6dfb68a53b6ae42bbb816dddf"],"score":34.78}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2.1","peer_count":82,"percentile":72.84,"percentile_exact":{"denominator":81,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","included":true,"model_id":"m-f892623396612d9b","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","included":true,"model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":78,"percentile":68.83,"percentile_exact":{"denominator":77,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","peer_count":24,"percentile":34.78,"percentile_exact":{"denominator":23,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35}],"score":64.85,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":25250},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2000},"family":"FrontierCode","row_ids":["4086551829a96ee8ea164e44d1e7c1c5"],"score":40.82},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":500},"family":"Vals Code Migration","row_ids":["68fce01452fe549878e8933d04edaa05"],"score":38.46}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","included":true,"model_id":"m-f892623396612d9b","original_version_label":"2.0","peer_count":50,"percentile":40.82,"percentile_exact":{"denominator":49,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","included":true,"model_id":"m-f892623396612d9b","original_version_label":"1.1","peer_count":27,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0}],"score":39.64,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":62200},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1200},"family":"GDP.pdf","row_ids":["32f9ed0b80b3e83e515a9f3fb34bc9b3"],"score":63.16},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3200},"family":"AA-Omniscience","row_ids":["ec9021441721b216b9626f97b96062f1","880d1169acb398e66460aed2eb721de0"],"score":60.38}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072}],"score":61.77,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":650},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":650},"family":"MMMU-Pro","row_ids":["e0d9cab1ce675bebaf2c47a0db9ddea9"],"score":38.24}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902}],"score":38.24,"weight":10}],"eligible":true,"id":"m-f892623396612d9b","in_window":true,"name":"Step 5 Preview","official_evidence":{"artifact_sha256":"f316d16c93b2d4ce98e733917634947e32a9218037b9e1d3d427de4b75275ddb","basis":"official_release_page","date":"2026-09-20","sha256":"f316d16c93b2d4ce98e733917634947e32a9218037b9e1d3d427de4b75275ddb","url":"https://www.stepfun.com/news/step-5-preview"},"primary_profile_id":"native","profile_id":"native","rank":37,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f80ea7a9c8b1ad0f866390a070dec6ae","included":true,"model_id":"m-f892623396612d9b","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row33"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row33","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ea28428a53ddb22639d7f5b7e50547a2","included":true,"model_id":"m-f892623396612d9b","original_version_label":"Diamond","peer_count":64,"percentile":71.43,"percentile_exact":{"denominator":7,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.243,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"step-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.243},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5e83ad92c83f666dbcd3cdc392d63b6","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2 Hard","peer_count":52,"percentile":59.8,"percentile_exact":{"denominator":51,"numerator":3050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.325,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"step-5:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.325},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"519c33c19b165c6197391af117e522e3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.379793510324484,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"step-5:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379793510324484},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d52704f212d466a4d653d53f1794d1db","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v2.1","peer_count":82,"percentile":72.84,"percentile_exact":{"denominator":81,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1593.99,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"step-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1593.99},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Step 5 Preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"d5a9a67230f3e72676fe2415a004dd47","included":true,"model_id":"m-f892623396612d9b","original_version_label":"4.0","peer_count":49,"percentile":57.29,"percentile_exact":{"denominator":24,"numerator":1375},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.82,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.82%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.82},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"85e95115c24e6ebd5d7a72360c176406","included":true,"model_id":"m-f892623396612d9b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5095659419096288,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"step-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5095659419096288},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5342274720726a814f5f5d822f07b510","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":78,"percentile":68.83,"percentile_exact":{"denominator":77,"numerator":5300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1425.18,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"step-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1425.18},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6be4e7acd5ffa2c31291fe47e51a2a60","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA run","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4717994628469114,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"step-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4717994628469114},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1f372ec6dfb68a53b6ae42bbb816dddf","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA pass^5","peer_count":24,"percentile":34.78,"percentile_exact":{"denominator":23,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.35,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"step-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.35},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"4086551829a96ee8ea164e44d1e7c1c5","included":true,"model_id":"m-f892623396612d9b","original_version_label":"2.0","peer_count":50,"percentile":40.82,"percentile_exact":{"denominator":49,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.120010101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"step-5:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.120010101010101},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Step 5 Preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"68fce01452fe549878e8933d04edaa05","included":true,"model_id":"m-f892623396612d9b","original_version_label":"1.1","peer_count":27,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":47.0,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"47.00%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row17"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row17","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":47.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"32f9ed0b80b3e83e515a9f3fb34bc9b3","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current AA protocol","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.148,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"step-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9021441721b216b9626f97b96062f1","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":58.49,"percentile_exact":{"denominator":53,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3636666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3636666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Step 5 Preview","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"880d1169acb398e66460aed2eb721de0","included":true,"model_id":"m-f892623396612d9b","original_version_label":"v1.1","peer_count":54,"percentile":62.26,"percentile_exact":{"denominator":53,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6867469879518072,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"step-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6867469879518072},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Step 5 Preview","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"e0d9cab1ce675bebaf2c47a0db9ddea9","included":true,"model_id":"m-f892623396612d9b","original_version_label":"current Standard 10","peer_count":35,"percentile":38.24,"percentile_exact":{"denominator":17,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763583815028902,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"step-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"step-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763583815028902}],"release_date":"2026-09-20","release_status":"published","release_verified":true,"result_updated_at":null,"score":52.79,"score_exact":{"denominator":376374110112,"numerator":19867815404455},"source_id":"stepfun/step-5-preview"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":3591,"numerator":160750},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":350},"family":"Terminal-Bench-Science","row_ids":["17afba9dea11c1bfb4fcda8f7ceb8fe4"],"score":18.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3100},"family":"HLE","row_ids":["2e68084b0515fdc6a19fe2b7a59814d2"],"score":49.21},{"benchmark_version":"2 Hard","exact":{"denominator":3,"numerator":200},"family":"CritPt","row_ids":["2f0eebfa718074ef0cab557ff07d209b"],"score":66.67}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","peer_count":64,"percentile":49.21,"percentile_exact":{"denominator":63,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","peer_count":52,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34}],"score":44.76,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":605682,"numerator":40432325},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":800},"family":"GDPval-AA","row_ids":["82bb5fcac208ae67d3e4a7d3de124d5d"],"score":88.89},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":325},"family":"Terminal-Bench","row_ids":["65da79b527ae8b650ca14c9d0ac7489d"],"score":54.17},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1375},"family":"AutomationBench","row_ids":["45f30ef9c9577f4abce574fa0b12543a"],"score":72.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":6100},"family":"AA-Briefcase","row_ids":["c6fb2c23c770d5e7bda4e380c2517fe7"],"score":79.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":900},"family":"AA-AnalystAgent","row_ids":["59ce77c04181c398c36c852841fb4ecf"],"score":39.13}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","peer_count":82,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","peer_count":49,"percentile":54.17,"percentile_exact":{"denominator":6,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":78,"percentile":79.22,"percentile_exact":{"denominator":77,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","peer_count":24,"percentile":39.13,"percentile_exact":{"denominator":23,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375}],"score":66.76,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":34275},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3200},"family":"FrontierCode","row_ids":["f7aceff949a2ebec04ba4fe1d1d7e419"],"score":65.31},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":550},"family":"Vals Code Migration","row_ids":["c811b7493c2ca32eedcc762a5f354d58"],"score":42.31}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","peer_count":50,"percentile":65.31,"percentile_exact":{"denominator":49,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","peer_count":27,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79}],"score":53.81,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":58000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1400},"family":"GDP.pdf","row_ids":["0c01c6860fffc32ef81ab819ece21423"],"score":73.68},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2200},"family":"AA-Omniscience","row_ids":["d99f9179103c8156a9cedc7180b05c42","b3118336d86ca34d1fd6619d5a595882"],"score":41.51}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376}],"score":57.6,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":200},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["73994f722364016b18f62a10e6695b38"],"score":22.22}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","peer_count":28,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2}],"score":22.22,"weight":10}],"eligible":true,"id":"m-dabee5e3d87d8a7a","in_window":true,"name":"MiMo-V2.6-Pro","official_evidence":{"artifact_sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","basis":"official_release_page","date":"2026-09-22","sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","url":"https://mimo.mi.com/docs/en-US/updates/model"},"primary_profile_id":"native","profile_id":"native","rank":38,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"17afba9dea11c1bfb4fcda8f7ceb8fe4","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"0.1.0","peer_count":39,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":2.86,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"2.86%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":2.86},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2e68084b0515fdc6a19fe2b7a59814d2","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"Diamond","peer_count":64,"percentile":49.21,"percentile_exact":{"denominator":63,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.152,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f0eebfa718074ef0cab557ff07d209b","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2 Hard","peer_count":52,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.34,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"82bb5fcac208ae67d3e4a7d3de124d5d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v2.1","peer_count":82,"percentile":88.89,"percentile_exact":{"denominator":9,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1689.4,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1689.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"65da79b527ae8b650ca14c9d0ac7489d","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"4.0","peer_count":49,"percentile":54.17,"percentile_exact":{"denominator":6,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":31.31,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"31.31%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row23","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":31.31},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"45f30ef9c9577f4abce574fa0b12543a","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5862340943419085,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5862340943419085},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c6fb2c23c770d5e7bda4e380c2517fe7","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":78,"percentile":79.22,"percentile_exact":{"denominator":77,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1511.88,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1511.88},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"59ce77c04181c398c36c852841fb4ecf","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA pass^5","peer_count":24,"percentile":39.13,"percentile_exact":{"denominator":23,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mimo-v2-6-pro:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"f7aceff949a2ebec04ba4fe1d1d7e419","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"2.0","peer_count":50,"percentile":65.31,"percentile_exact":{"denominator":49,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.239701262626263,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.239701262626263},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"c811b7493c2ca32eedcc762a5f354d58","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"1.1","peer_count":27,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":48.79,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"48.79%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row16"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row16","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":48.79},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0c01c6860fffc32ef81ab819ece21423","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current AA protocol","peer_count":77,"percentile":73.68,"percentile_exact":{"denominator":19,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.192,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.192},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d99f9179103c8156a9cedc7180b05c42","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":56.6,"percentile_exact":{"denominator":53,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.348,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.348},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Pro","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b3118336d86ca34d1fd6619d5a595882","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.39775051124744376,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.39775051124744376},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Pro (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"73994f722364016b18f62a10e6695b38","included":true,"model_id":"m-dabee5e3d87d8a7a","original_version_label":"current public protocol","peer_count":28,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.2,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"6e073bffa650bae01b4fe7e7"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"6e073bffa650bae01b4fe7e7","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.2}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":52.38,"score_exact":{"denominator":35054451432,"numerator":1836253825765},"source_id":"xiaomi/mimo-v2.6-pro"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3717,"numerator":175250},"families":[{"benchmark_version":"2026","exact":{"denominator":1,"numerator":100},"family":"AIME","row_ids":["0f70d96aa863b344234e9f6fb914e9f2"],"score":100.0},{"benchmark_version":"500-item subset","exact":{"denominator":3,"numerator":200},"family":"MATH-500","row_ids":["1f0d5271a4aa3a380367a096ae7ae210"],"score":66.67},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1200},"family":"GPQA Diamond","row_ids":["d95e6a6f385be987c8949ca6952efa66"],"score":20.34},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":100},"family":"HLE","row_ids":["75182b4b31b9383c53f3f18d0e09c65d"],"score":1.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"2026","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":20.34,"percentile_exact":{"denominator":59,"numerator":1200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond","peer_count":64,"percentile":1.59,"percentile_exact":{"denominator":63,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038}],"score":47.15,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":18486468,"numerator":627714865},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":1225},"family":"tau3-Banking","row_ids":["dc8c6983f496e505270c3f1d2e08a1c1"],"score":47.12},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1600},"family":"GDPval-AA","row_ids":["f2ca1c75e7e337b37337e6b5eb3f5cd5"],"score":19.75},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["49929d3422a54a6833b2d573268b4f8f"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":225},"family":"AutomationBench","row_ids":["4a74a50d45014989b724a431c32f6c56"],"score":11.84},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":900},"family":"AA-Briefcase","row_ids":["8a20412103b559527bf3c1a6c26511e5"],"score":11.69},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["9fed535a16c1cf8e07c033c97e3a6783"],"score":100.0}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","included":true,"model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v2.1","peer_count":82,"percentile":19.75,"percentile_exact":{"denominator":81,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","included":true,"model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v1.1","peer_count":78,"percentile":11.69,"percentile_exact":{"denominator":77,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v4","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6}],"score":33.96,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":100},"families":[{"benchmark_version":"release_v6","exact":{"denominator":1,"numerator":100},"family":"LiveCodeBench","row_ids":["97029c45627dc4bbce9101042fef65e2"],"score":100.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1}],"score":100.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":171,"numerator":5495},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"IFBench","row_ids":["30e5129b448d30519f98acdcb4661feb"],"score":44.44},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":100},"family":"Multi-IF","row_ids":["ca424296b6739f797da253012aa25cc9"],"score":33.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":150},"family":"GDP.pdf","row_ids":["9ffb2abc9760a5a8e507be89d245bb8f"],"score":7.89},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"IFEval","row_ids":["c5c1fcc77312c8bfdb8f9a2733828c50"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":75},"family":"MMLU-Pro","row_ids":["fdfed7f2ef133f7ec35f427f9ee8f813"],"score":75.0}],"family_count":5,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8}],"score":32.13,"weight":20}],"eligible":true,"id":"m-407d9206347b790d","in_window":true,"name":"MiniCPM5-2B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-07","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":39,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0f70d96aa863b344234e9f6fb914e9f2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"2026","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":86.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1f0d5271a4aa3a380367a096ae7ae210","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":94.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"94.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":94.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d95e6a6f385be987c8949ca6952efa66","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":20.34,"percentile_exact":{"denominator":59,"numerator":1200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.702020202020202,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm5-2b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.702020202020202},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"75182b4b31b9383c53f3f18d0e09c65d","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Diamond","peer_count":64,"percentile":1.59,"percentile_exact":{"denominator":63,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.038,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minicpm5-2b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dc8c6983f496e505270c3f1d2e08a1c1","included":true,"model_id":"m-407d9206347b790d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minicpm5-2b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f2ca1c75e7e337b37337e6b5eb3f5cd5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v2.1","peer_count":82,"percentile":19.75,"percentile_exact":{"denominator":81,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":714.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm5-2b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":714.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49929d3422a54a6833b2d573268b4f8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"minicpm5-2b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a74a50d45014989b724a431c32f6c56","included":true,"model_id":"m-407d9206347b790d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018006007862431707,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minicpm5-2b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018006007862431707},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8a20412103b559527bf3c1a6c26511e5","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v1.1","peer_count":78,"percentile":11.69,"percentile_exact":{"denominator":77,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":423.29,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minicpm5-2b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":423.29},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9fed535a16c1cf8e07c033c97e3a6783","included":true,"model_id":"m-407d9206347b790d","original_version_label":"v4","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":66.6,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.6},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"97029c45627dc4bbce9101042fef65e2","included":true,"model_id":"m-407d9206347b790d","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":69.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"69.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":69.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"30e5129b448d30519f98acdcb4661feb","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":66.3,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"66.3","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":66.3},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ca424296b6739f797da253012aa25cc9","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.8,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.8},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ffb2abc9760a5a8e507be89d245bb8f","included":true,"model_id":"m-407d9206347b790d","original_version_label":"current AA protocol","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.014,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm5-2b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minicpm5-2b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c5c1fcc77312c8bfdb8f9a2733828c50","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current original Google dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":86.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"86.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":86.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"MiniCPM5-2B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"fdfed7f2ef133f7ec35f427f9ee8f813","included":true,"model_id":"m-407d9206347b790d","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column MiniCPM5-2B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":70.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"70.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column1","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":70.8}],"release_date":"2026-09-07","release_status":"published","release_verified":true,"result_updated_at":null,"score":51.89,"score_exact":{"denominator":19632629016,"numerator":1018776441535},"source_id":"openbmb/minicpm5-2b"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":6571656,"numerator":314311525},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4750},"family":"GPQA Diamond","row_ids":["ba847bf9e90b2e6c2d3a1b021f6f1ff2"],"score":80.51},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["2f71320c4fe6daaec73fc9a09a2e75a6"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":3200},"family":"HLE","row_ids":["ee3f7405021072ff698c63b6e1fa510f"],"score":50.79},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":350},"family":"CritPt","row_ids":["a3193e796745f7cc574aae69daceb5f1"],"score":20.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","included":true,"model_id":"m-c947385433db3f03","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond","peer_count":64,"percentile":50.79,"percentile_exact":{"denominator":63,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04}],"score":47.83,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":220077,"numerator":14695565},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1250},"family":"tau3-Banking","row_ids":["02b14b2382923c4835faa17152c2d94b"],"score":96.15},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1700},"family":"GDPval-AA","row_ids":["022f63e78e1f4658e8af4b434e595fa3"],"score":62.96},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":160},"family":"Terminal-Bench","row_ids":["bb15f242cbb67b0a52ecb1efaeaf802e"],"score":53.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1100},"family":"AutomationBench","row_ids":["07fee1c32e6f2287a67ddf683ea02c91"],"score":57.89},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":700},"family":"AA-Briefcase","row_ids":["bee797f9fafa487f710b7abf6c3fddcf"],"score":63.64},{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"EnterpriseOps-Gym","row_ids":["cea12af74b5609df87befba4292e8ab2"],"score":66.67}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","included":true,"model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2.1","peer_count":82,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","included":true,"model_id":"m-c947385433db3f03","original_version_label":"4.0","peer_count":76,"percentile":53.33,"percentile_exact":{"denominator":3,"numerator":160},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","included":true,"model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":78,"percentile":63.64,"percentile_exact":{"denominator":11,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA run","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471}],"score":66.77,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":21,"numerator":800},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":300},"family":"FrontierCode","row_ids":["328d104cb62d50eef3ae2e92cc4db522"],"score":42.86},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":100},"family":"DeepSWE","row_ids":["b993c11069821aca78f26acbce738dbc"],"score":33.33}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","included":true,"model_id":"m-c947385433db3f03","original_version_label":"2.0","peer_count":50,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","included":true,"model_id":"m-c947385433db3f03","original_version_label":"1.1","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2}],"score":38.1,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":9063,"numerator":536350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"IFBench","row_ids":["860832a18411b625ebef7842433113ca"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1300},"family":"GDP.pdf","row_ids":["4189c00054cffddacad0958de8d0f156"],"score":68.42},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["5ce6a4a59546a3ac7bd2230a498f7f48","ee80bfee31969adf86abcbeaae10d4e4"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current published protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742}],"score":59.18,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":600},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":600},"family":"MMMU-Pro","row_ids":["fa359632aae944ad7f900cc61aa55351"],"score":35.29}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821}],"score":35.29,"weight":10}],"eligible":true,"id":"m-c947385433db3f03","in_window":true,"name":"Qwen3.8 27B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-14","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":40,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ba847bf9e90b2e6c2d3a1b021f6f1ff2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":80.51,"percentile_exact":{"denominator":59,"numerator":4750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.905050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-8-27b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.905050505050505},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f71320c4fe6daaec73fc9a09a2e75a6","included":true,"model_id":"m-c947385433db3f03","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"qwen3-8-27b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ee3f7405021072ff698c63b6e1fa510f","included":true,"model_id":"m-c947385433db3f03","original_version_label":"Diamond","peer_count":64,"percentile":50.79,"percentile_exact":{"denominator":63,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.153,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-8-27b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.153},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a3193e796745f7cc574aae69daceb5f1","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"qwen3-8-27b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"02b14b2382923c4835faa17152c2d94b","included":true,"model_id":"m-c947385433db3f03","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":96.15,"percentile_exact":{"denominator":13,"numerator":1250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.480412371134021,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-8-27b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.480412371134021},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"022f63e78e1f4658e8af4b434e595fa3","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v2.1","peer_count":82,"percentile":62.96,"percentile_exact":{"denominator":27,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1426.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-8-27b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1426.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bb15f242cbb67b0a52ecb1efaeaf802e","included":true,"model_id":"m-c947385433db3f03","original_version_label":"4.0","peer_count":76,"percentile":53.33,"percentile_exact":{"denominator":3,"numerator":160},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0555555555555556,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-8-27b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0555555555555556},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"07fee1c32e6f2287a67ddf683ea02c91","included":true,"model_id":"m-c947385433db3f03","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.48242443682241926,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-8-27b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.48242443682241926},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bee797f9fafa487f710b7abf6c3fddcf","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":78,"percentile":63.64,"percentile_exact":{"denominator":11,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1391.36,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-8-27b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1391.36},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cea12af74b5609df87befba4292e8ab2","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA run","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4422560429722471,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-8-27b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4422560429722471},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"328d104cb62d50eef3ae2e92cc4db522","included":true,"model_id":"m-c947385433db3f03","original_version_label":"2.0","peer_count":50,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149142929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"qwen3-8-27b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149142929292929},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"b993c11069821aca78f26acbce738dbc","included":true,"model_id":"m-c947385433db3f03","original_version_label":"1.1","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":42.2,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"42.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":42.2},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.8-27B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"860832a18411b625ebef7842433113ca","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current published protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.5,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column2","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.5},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4189c00054cffddacad0958de8d0f156","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current AA protocol","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.164,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-8-27b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.164},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ce6a4a59546a3ac7bd2230a498f7f48","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.173,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Qwen3.8 27B (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ee80bfee31969adf86abcbeaae10d4e4","included":true,"model_id":"m-c947385433db3f03","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7138250705360742,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"qwen3-8-27b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7138250705360742},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.8 27B (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"fa359632aae944ad7f900cc61aa55351","included":true,"model_id":"m-c947385433db3f03","original_version_label":"current Standard 10","peer_count":35,"percentile":35.29,"percentile_exact":{"denominator":17,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.763005780346821,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-8-27b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-8-27b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.763005780346821}],"release_date":"2026-08-14","release_status":"published","release_verified":true,"result_updated_at":"2026-08-26","score":51.64,"score_exact":{"denominator":2620592406432,"numerator":135314940818345},"source_id":"alibaba/qwen3.8-27b"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":164300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2500},"family":"GPQA Diamond","row_ids":["4ac219e06b407a4da2953a84f21a360f"],"score":42.37},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2900},"family":"HLE","row_ids":["5706c727b7c4831ffe377f7ebcc17796"],"score":46.03}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":42.37,"percentile_exact":{"denominator":59,"numerator":2500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","peer_count":64,"percentile":46.03,"percentile_exact":{"denominator":63,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146}],"score":44.2,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":293031139},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":250},"family":"tau3-Banking","row_ids":["c4bd6d87d676b473fe7bbadf8b7238e4"],"score":19.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1100},"family":"GDPval-AA","row_ids":["0bf079be27f6b7285cbdd993ff5b5a93"],"score":13.58},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":178},"family":"Terminal-Bench","row_ids":["4473f9853fd2b2892365909edf9631fb"],"score":59.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":975},"family":"AutomationBench","row_ids":["20e66aebafb2226fab822f3a7524d54c"],"score":51.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3600},"family":"AA-Briefcase","row_ids":["110052f3248350cfda8c6580116d3dd9"],"score":46.75}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","peer_count":82,"percentile":13.58,"percentile_exact":{"denominator":81,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"4.0","peer_count":76,"percentile":59.33,"percentile_exact":{"denominator":3,"numerator":178},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":51.32,"percentile_exact":{"denominator":19,"numerator":975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","peer_count":78,"percentile":46.75,"percentile_exact":{"denominator":77,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66}],"score":38.04,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":1450},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1450},"family":"GDP.pdf","row_ids":["9ab89470db83472d1b5d2ed28ebc6783"],"score":76.32}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202}],"score":76.32,"weight":20}],"eligible":true,"id":"m-3fe24456aa661a33","in_window":true,"name":"GPT-5.5 Instant (June 2026)","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":41,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4ac219e06b407a4da2953a84f21a360f","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":42.37,"percentile_exact":{"denominator":59,"numerator":2500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.823232323232323,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gpt-5-5-instant-06-26:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.823232323232323},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5706c727b7c4831ffe377f7ebcc17796","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"Diamond","peer_count":64,"percentile":46.03,"percentile_exact":{"denominator":63,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.146,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-5-5-instant-06-26:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.146},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bd6d87d676b473fe7bbadf8b7238e4","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.117525773195876,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gpt-5-5-instant-06-26:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.117525773195876},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0bf079be27f6b7285cbdd993ff5b5a93","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v2.1","peer_count":82,"percentile":13.58,"percentile_exact":{"denominator":81,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":516.27,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-5-5-instant-06-26:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":516.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4473f9853fd2b2892365909edf9631fb","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"4.0","peer_count":76,"percentile":59.33,"percentile_exact":{"denominator":3,"numerator":178},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.126262626262626,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gpt-5-5-instant-06-26:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.126262626262626},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20e66aebafb2226fab822f3a7524d54c","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":51.32,"percentile_exact":{"denominator":19,"numerator":975},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.36436554387118747,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-5-5-instant-06-26:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.36436554387118747},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"110052f3248350cfda8c6580116d3dd9","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"v1.1","peer_count":78,"percentile":46.75,"percentile_exact":{"denominator":77,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1086.66,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-5-5-instant-06-26:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1086.66},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-5.5 Instant (June 2026)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9ab89470db83472d1b5d2ed28ebc6783","included":true,"model_id":"m-3fe24456aa661a33","original_version_label":"current AA protocol","peer_count":77,"percentile":76.32,"percentile_exact":{"denominator":19,"numerator":1450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.202,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-5-5-instant-06-26:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-5-5-instant-06-26:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202}],"release_date":"2026-06-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":51.18,"score_exact":{"denominator":1272485214,"numerator":65122894901},"source_id":"published/aa:gpt-5-5-instant-06-26"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":800394,"numerator":45089825},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5350},"family":"GPQA Diamond","row_ids":["a4b58fb51111dbcc6b81f75ff3f7fc61"],"score":90.68},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["0898509511ccb710480bf98d3742bbb0"],"score":31.58},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1300},"family":"HLE","row_ids":["d543f8764a78abfabdc05d4f312b8b2f"],"score":61.9},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":700},"family":"CritPt","row_ids":["a2d7db240ac12326c1a9bfd85ff9b349"],"score":41.18}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","peer_count":64,"percentile":61.9,"percentile_exact":{"denominator":21,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","peer_count":52,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235}],"score":56.33,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":145236,"numerator":8347475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["76dde74584cddc5a2d8618786061726c"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1025},"family":"tau3-Banking","row_ids":["20bfc5477d34134d5c7c99539fd481a0"],"score":78.85},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":200},"family":"GDPval-AA","row_ids":["727c30106a7ec7d9e1c1205d514e6937"],"score":66.67},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":125},"family":"Terminal-Bench","row_ids":["35687651705594abdbf62890a0bd5a87"],"score":31.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1300},"family":"AutomationBench","row_ids":["63186f8028fad18deb2fa4c75b5a0888"],"score":68.42},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":400},"family":"AA-Briefcase","row_ids":["40fe84ac7f2d6389219a5c06f59da083"],"score":57.14},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["6c40f85770ab1fe03b45de2938f913b3"],"score":100.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":78.85,"percentile_exact":{"denominator":13,"numerator":1025},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","peer_count":82,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"4.0","peer_count":49,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":78,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731}],"score":57.48,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":441,"numerator":17000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2400},"family":"FrontierCode","row_ids":["2e337d0b2b65e72e819796e03cdc53e7"],"score":48.98},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"Vals Code Migration","row_ids":["7e542189106406c0f4ba769f7753c91e"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":50},"family":"DeepSWE","row_ids":["63840894bc67cdd6c5a9cd6e26e8b9d7"],"score":16.67}],"family_count":3,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"2.0","peer_count":50,"percentile":48.98,"percentile_exact":{"denominator":49,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":27,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7}],"score":38.55,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2485600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1000},"family":"GDP.pdf","row_ids":["c287cda1f48a055602b23132b6d39858"],"score":52.63},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":900},"family":"SimpleQA Verified","row_ids":["665cc51403349d6e59eed00558808309"],"score":52.94},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["3f8c786b8fc874f5d0f0ae1174917198","775332c591df3bb4219419f088bb5ff5"],"score":39.62}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429}],"score":48.4,"weight":20}],"eligible":true,"id":"m-f936349cbdb1def9","in_window":true,"name":"DeepSeek V4 Pro 0813","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-13","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":42,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a4b58fb51111dbcc6b81f75ff3f7fc61","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":90.68,"percentile_exact":{"denominator":59,"numerator":5350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.928282828282828,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"deepseek-v4-pro:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.928282828282828},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"0898509511ccb710480bf98d3742bbb0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row31"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row31","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d543f8764a78abfabdc05d4f312b8b2f","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Diamond","peer_count":64,"percentile":61.9,"percentile_exact":{"denominator":21,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.195,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"deepseek-v4-pro:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2d7db240ac12326c1a9bfd85ff9b349","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2 Hard","peer_count":52,"percentile":41.18,"percentile_exact":{"denominator":17,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.235,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"deepseek-v4-pro:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / official Hy4 comparison, APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"76dde74584cddc5a2d8618786061726c","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"Current published protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":32.4,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"hy4-image-apex-comparator:DeepSeek V4 Pro 0813","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"20bfc5477d34134d5c7c99539fd481a0","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":78.85,"percentile_exact":{"denominator":13,"numerator":1025},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.395876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"deepseek-v4-pro:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.395876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"727c30106a7ec7d9e1c1205d514e6937","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v2.1","peer_count":82,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1457.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"deepseek-v4-pro:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1457.37},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"35687651705594abdbf62890a0bd5a87","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"4.0","peer_count":49,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":14.14,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"14.14%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":14.14},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"63186f8028fad18deb2fa4c75b5a0888","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":68.42,"percentile_exact":{"denominator":19,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5671165546344002,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"deepseek-v4-pro:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5671165546344002},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40fe84ac7f2d6389219a5c06f59da083","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":78,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1253.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"deepseek-v4-pro:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1253.16},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6c40f85770ab1fe03b45de2938f913b3","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA run","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4959713518352731,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"deepseek-v4-pro:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4959713518352731},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"2e337d0b2b65e72e819796e03cdc53e7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"2.0","peer_count":50,"percentile":48.98,"percentile_exact":{"denominator":49,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.170742929292929,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"deepseek-v4-pro:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.170742929292929},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"7e542189106406c0f4ba769f7753c91e","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":27,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":51.99,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"51.99%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row14"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row14","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":51.99},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Pro","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"63840894bc67cdd6c5a9cd6e26e8b9d7","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"1.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":62.7,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"62.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column5"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column5","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":62.7},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c287cda1f48a055602b23132b6d39858","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"current AA protocol","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.114,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"deepseek-v4-pro:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.114},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-pro-0813_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"665cc51403349d6e59eed00558808309","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"run 1.2.0","peer_count":35,"percentile":52.94,"percentile_exact":{"denominator":17,"numerator":900},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5290581162324649,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:20.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8cbbda92237229d94a9b20c0"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"8cbbda92237229d94a9b20c0","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5290581162324649},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f8c786b8fc874f5d0f0ae1174917198","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.452,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.452},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"DeepSeek V4 Pro 0813 (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"775332c591df3bb4219419f088bb5ff5","included":true,"model_id":"m-f936349cbdb1def9","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.05900243309002429,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"deepseek-v4-pro:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"deepseek-v4-pro:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05900243309002429}],"release_date":"2026-08-13","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":50.94,"score_exact":{"denominator":416912428296,"numerator":21235615018525},"source_id":"deepseek/deepseek-v4-pro-0813"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":27846,"numerator":1152575},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":50},"family":"FrontierMath","row_ids":["8462c2a29fbbd735e478ce804db0a935"],"score":5.56},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["9bdac139a55bab7bdafc52c1059a96f7"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":400},"family":"HLE","row_ids":["3c62c893b3f2b3d94df10089539fb20a"],"score":57.14},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2500},"family":"CritPt","row_ids":["e52491cdc38740f220e3951458e0d11e"],"score":49.02}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","peer_count":64,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","peer_count":52,"percentile":49.02,"percentile_exact":{"denominator":51,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27}],"score":41.39,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":13167,"numerator":878050},"families":[{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":200},"family":"Terminal-Bench","row_ids":["038de3c11daf5dbf02e5d1c9dd11ccd7"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1350},"family":"AutomationBench","row_ids":["7fa3942a02994b01f2c4e2365b0583bc"],"score":71.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4800},"family":"AA-Briefcase","row_ids":["bc12c3c00877323848c26ea730d935cc"],"score":62.34}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"4.0","peer_count":76,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":78,"percentile":62.34,"percentile_exact":{"denominator":77,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31}],"score":66.69,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":2200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2200},"family":"FrontierCode","row_ids":["c316e6a1765e379f023423782e08477b"],"score":44.9}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"2.0","peer_count":50,"percentile":44.9,"percentile_exact":{"denominator":49,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313}],"score":44.9,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2154950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1350},"family":"GDP.pdf","row_ids":["a6443ea346efcda5f00f1ce182e0d982"],"score":71.05},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":50},"family":"SimpleQA Verified","row_ids":["39e767a0ed9e71374bab635fd9e2ce92"],"score":2.94},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2750},"family":"AA-Omniscience","row_ids":["6656d06e85a16beb97db67e4960f17c9","8175922707070fb09367c5a6424e1c3a"],"score":51.89}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018}],"score":41.96,"weight":20}],"eligible":true,"id":"m-72b806dc67894cf3","in_window":true,"name":"Mistral Large 4 Preview","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":43,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8462c2a29fbbd735e478ce804db0a935","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-12","result_timestamp":"2026-10-12T21:38:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cd363f32842485ea3c5ba372"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"cd363f32842485ea3c5ba372","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9bdac139a55bab7bdafc52c1059a96f7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c62c893b3f2b3d94df10089539fb20a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"Diamond","peer_count":64,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.173,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.173},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e52491cdc38740f220e3951458e0d11e","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v2 Hard","peer_count":52,"percentile":49.02,"percentile_exact":{"denominator":51,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.27,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.27},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"038de3c11daf5dbf02e5d1c9dd11ccd7","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"4.0","peer_count":76,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.247474747474747,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.247474747474747},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fa3942a02994b01f2c4e2365b0583bc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5783010541628066,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5783010541628066},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bc12c3c00877323848c26ea730d935cc","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":78,"percentile":62.34,"percentile_exact":{"denominator":77,"numerator":4800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1348.31,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1348.31},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c316e6a1765e379f023423782e08477b","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"2.0","peer_count":50,"percentile":44.9,"percentile_exact":{"denominator":49,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.149581313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.149581313131313},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"a6443ea346efcda5f00f1ce182e0d982","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"current AA protocol","peer_count":77,"percentile":71.05,"percentile_exact":{"denominator":19,"numerator":1350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.176,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.176},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"mistral-large-4-0_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"39e767a0ed9e71374bab635fd9e2ce92","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"run 1.2.0","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.2,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-09","result_timestamp":"2026-10-09T14:12:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d10ea299849c4e837eef669"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1d10ea299849c4e837eef669","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6656d06e85a16beb97db67e4960f17c9","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.26,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Large 4 (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8175922707070fb09367c5a6424e1c3a","included":true,"model_id":"m-72b806dc67894cf3","original_version_label":"v1.1","peer_count":54,"percentile":73.58,"percentile_exact":{"denominator":53,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.718018018018018,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.718018018018018}],"release_date":"2026-10-06","release_status":"published","release_verified":true,"result_updated_at":"2026-10-12","score":49.32,"score_exact":{"denominator":38864717892,"numerator":1916929064525},"source_id":"mistral/mistral-large-4"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":513,"numerator":7325},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":50},"family":"FrontierMath","row_ids":["3121b8aa930fa12d352b2910814de607"],"score":5.56},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":100},"family":"GPQA Diamond","row_ids":["6c2c7071889205aa8d51f6bb7f06fb84"],"score":33.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["90672f02fa6c8adab2a2801b7c2aedb3"],"score":3.95}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"score":14.28,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":72,"numerator":5375},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":325},"family":"Terminal-Bench","row_ids":["4ce54d74c2045140751a930d5f81e63e"],"score":40.63},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["e28fafc89bbfaddac23f5b76027e9f84"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":250},"family":"AA-Briefcase","row_ids":["59421edaa49a1a1c8543f29553d821c6"],"score":83.33}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","peer_count":49,"percentile":40.63,"percentile_exact":{"denominator":8,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855}],"score":74.65,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":1100},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":900},"family":"Vals Code Migration","row_ids":["3beb1575e4852b44427673ea326ae656"],"score":69.23},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"DeepSWE","row_ids":["3bd3c3d2983a2fb1db83fc1872674955"],"score":100.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":27,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2}],"score":84.62,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":550},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":550},"family":"SimpleQA Verified","row_ids":["66b5a46f01ac8af0a7586871cb050ba1"],"score":32.35}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429}],"score":32.35,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"Chartography","row_ids":["e5a64d88a17b8e69890219a69e61c957"],"score":33.33}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9}],"score":33.33,"weight":10}],"eligible":true,"id":"m-99183a5b3b302b0d","in_window":true,"name":"DeepSeek V4.1 Flash","official_evidence":{"artifact_sha256":"df2d95745d8bc7620e1be8267ede37334adde98086da78ba1e49574f407a619a","basis":"official_release_page","date":"2026-09-10","sha256":"df2d95745d8bc7620e1be8267ede37334adde98086da78ba1e49574f407a619a","url":"https://www.deepseek.com/en/news/deepseek-v4-1-flash/"},"primary_profile_id":"native","profile_id":"native","rank":44,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"3121b8aa930fa12d352b2910814de607","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.2682926829268293,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-08","result_timestamp":"2026-10-08T19:02:32.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"bd8c94b9074fabe3952908a3"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"bd8c94b9074fabe3952908a3","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.2682926829268293},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"6c2c7071889205aa8d51f6bb7f06fb84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4.1-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":90.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"90.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":90.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"90672f02fa6c8adab2a2801b7c2aedb3","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"4ce54d74c2045140751a930d5f81e63e","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"4.0","peer_count":49,"percentile":40.63,"percentile_exact":{"denominator":8,"numerator":325},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":19.7,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"19.70%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row32"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row32","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":19.7},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e28fafc89bbfaddac23f5b76027e9f84","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published developer protocol","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":54.8,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"54.8","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":54.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"59421edaa49a1a1c8543f29553d821c6","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"v1.1","peer_count":19,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45194274028629855,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7257ecc2d2507fe3407c4157"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7257ecc2d2507fe3407c4157","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45194274028629855},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"DeepSeek V4.1 Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"3beb1575e4852b44427673ea326ae656","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":27,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":58.94,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"58.94%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row9"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row9","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":58.94},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"3bd3c3d2983a2fb1db83fc1872674955","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"1.1","peer_count":7,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row DeepSWE v1.1 (Resolved); preserve page footnotes and task-specific settings","protocol_id":"b229959842eb7c98b68856d4","raw_value":74.2,"reference_revision":"3897b0e9600a181d7706e139083bc59a8e365a1fc76c715c7a70659270a0fe0d","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"74.2","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row10:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE v1.1 (Resolved)","source_record_id":"deepseek-v41-flash:table1:row10:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":74.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4.1-flash_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"66b5a46f01ac8af0a7586871cb050ba1","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"run 1.2.0","peer_count":35,"percentile":32.35,"percentile_exact":{"denominator":17,"numerator":550},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.429,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-13","result_timestamp":"2026-10-13T12:59:33.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"49de275fd8a4c1e54ea39658"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"49de275fd8a4c1e54ea39658","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.429},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4.1-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"e5a64d88a17b8e69890219a69e61c957","included":true,"model_id":"m-99183a5b3b302b0d","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":78.9,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"78.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column7"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column7","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":78.9}],"release_date":"2026-09-10","release_status":"published","release_verified":true,"result_updated_at":"2026-10-13","score":48.96,"score_exact":{"denominator":3627936,"numerator":177623315},"source_id":"deepseek/deepseek-v4.1-flash"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":41.67,"evaluator_count":2,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["6e4fcbf667751f178fabfb4106350208"],"score":31.58},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":25},"family":"HLE","row_ids":["5a5d1e52735d8efc5b98d11b601491a8"],"score":25.0}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"Diamond","peer_count":17,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4}],"score":28.29,"weight":25},{"effective_weight_percent":41.67,"evaluator_count":2,"exact":{"denominator":16,"numerator":975},"families":[{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":175},"family":"Terminal-Bench","row_ids":["73167c475c92b2f6b68a011a68a996ab"],"score":21.88},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":100},"family":"AA-Briefcase","row_ids":["0619cb6c1df6def76a850c294e50de63"],"score":100.0}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"v1.1","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162}],"score":60.94,"weight":25},{"effective_weight_percent":16.67,"evaluator_count":1,"exact":{"denominator":27,"numerator":1900},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1900},"family":"Humanity Sixth Sense","row_ids":["fcb2b9f9c2c7859d3369fef86db555f4"],"score":70.37}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","peer_count":28,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4}],"score":70.37,"weight":10}],"eligible":true,"id":"m-c0580b01221ed260","in_window":true,"name":"Muse Spark 1.3","official_evidence":{"artifact_sha256":"b221b103503442df9d920dbb3ff2002c3c9c86755300e3f9635018eaea8c2e56","basis":"official_release_page","date":"2026-09-02","sha256":"b221b103503442df9d920dbb3ff2002c3c9c86755300e3f9635018eaea8c2e56","url":"https://research.meta.ai/blog/introducing-muse-spark-1-3"},"primary_profile_id":"native","profile_id":"native","rank":45,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"6e4fcbf667751f178fabfb4106350208","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row26","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-1.3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"5a5d1e52735d8efc5b98d11b601491a8","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"Diamond","peer_count":17,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":25.4,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"559a55dc7ca9cecd7b34b7f7"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"559a55dc7ca9cecd7b34b7f7","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":25.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"73167c475c92b2f6b68a011a68a996ab","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row43"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row43","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.3 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"0619cb6c1df6def76a850c294e50de63","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"v1.1","peer_count":19,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.5161616161616162,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4b2c652f6db527c7ffa499cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4b2c652f6db527c7ffa499cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.5161616161616162},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fcb2b9f9c2c7859d3369fef86db555f4","included":true,"model_id":"m-c0580b01221ed260","original_version_label":"current public protocol","peer_count":28,"percentile":70.37,"percentile_exact":{"denominator":27,"numerator":1900},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":37.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f6f9872634097ab9e008fcdb"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"f6f9872634097ab9e008fcdb","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":37.4}],"release_date":"2026-09-02","release_status":"published","release_verified":true,"result_updated_at":null,"score":48.91,"score_exact":{"denominator":98496,"numerator":4817075},"source_id":"meta/muse-spark-1.3"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":45.45,"evaluator_count":2,"exact":{"denominator":144,"numerator":6425},"families":[{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":275},"family":"Terminal-Bench","row_ids":["bd6d00a622ee8666fe845f21e2f0650f"],"score":11.46},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":700},"family":"AA-Briefcase","row_ids":["2a4003430f96eaca6cf7fbdac62ade55"],"score":77.78}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154}],"score":44.62,"weight":25},{"effective_weight_percent":36.36,"evaluator_count":1,"exact":{"denominator":17,"numerator":1200},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1200},"family":"SimpleQA Verified","row_ids":["aa61e58d244812473e1af5a8764b4e07"],"score":70.59}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844}],"score":70.59,"weight":20},{"effective_weight_percent":18.18,"evaluator_count":1,"exact":{"denominator":27,"numerator":400},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":400},"family":"Humanity Sixth Sense","row_ids":["b1617164f055d3d72ddaf46f34da89d8"],"score":14.81}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","peer_count":28,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5}],"score":14.81,"weight":10}],"eligible":true,"id":"m-2f22b07da3af863c","in_window":true,"name":"Muse Spark 1.2","official_evidence":{"artifact_sha256":"92a7e106c1f5836b322f76df67124f4f08c92c8dbe8cfa67e78698bd91fb844f","basis":"official_release_page","date":"2026-08-05","sha256":"92a7e106c1f5836b322f76df67124f4f08c92c8dbe8cfa67e78698bd91fb844f","url":"https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2"},"primary_profile_id":"native","profile_id":"native","rank":46,"records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Muse Spark 1.2 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bd6d00a622ee8666fe845f21e2f0650f","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row48"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row48","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Muse Spark 1.2 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"2a4003430f96eaca6cf7fbdac62ade55","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"v1.1","peer_count":19,"percentile":77.78,"percentile_exact":{"denominator":9,"numerator":700},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.45151515151515154,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7107ef5a8f365a6fdbc6918d"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"7107ef5a8f365a6fdbc6918d","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.45151515151515154},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"muse-spark-1.2_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"aa61e58d244812473e1af5a8764b4e07","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"run 1.2.0","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.6030150753768844,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:49.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"df2aba8d1cbcb18b99f102ef"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"df2aba8d1cbcb18b99f102ef","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.6030150753768844},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark 1.2 (xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"b1617164f055d3d72ddaf46f34da89d8","included":true,"model_id":"m-2f22b07da3af863c","original_version_label":"current public protocol","peer_count":28,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"2b3cc73047b5d58af4508acc"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"2b3cc73047b5d58af4508acc","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.5}],"release_date":"2026-08-05","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":48.64,"score_exact":{"denominator":80784,"numerator":3929575},"source_id":"meta/muse-spark-1.2"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":469404,"numerator":20872015},"families":[{"benchmark_version":"2","exact":{"denominator":9,"numerator":200},"family":"FrontierMath","row_ids":["c2edb36eb77a36c2b7e3b5065aafdd4c"],"score":22.22},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4000},"family":"GPQA Diamond","row_ids":["bf4b9ff10398e116bbd0ed6eb4a0a88d"],"score":67.8},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["356d1ac12dc91c2d1cadb615b7e3023b"],"score":46.15},{"benchmark_version":"Diamond","exact":{"denominator":4,"numerator":125},"family":"HLE","row_ids":["04518aae51e855cbdd25679b26d57f51"],"score":31.25},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["3eef34fe8a68f3f65cb23f07be80c5e4"],"score":54.9}],"family_count":5,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","included":true,"model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":67.8,"percentile_exact":{"denominator":59,"numerator":4000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","included":true,"model_id":"m-db208dec31031420","original_version_label":"0.1","peer_count":53,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":44.46,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":237006,"numerator":13945607},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5000},"family":"GDPval-AA","row_ids":["c140b9f2f3b8b047f022d22a6670f4c4"],"score":61.73},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":164},"family":"Terminal-Bench","row_ids":["6d1fe83c673cc3e53108a1f798f9857e"],"score":54.67},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1150},"family":"AutomationBench","row_ids":["cc297748bec17bb0b4aa16e861980849"],"score":60.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4500},"family":"AA-Briefcase","row_ids":["00f62480bdc3ab7dea00bbda96ff9331"],"score":58.44}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2.1","peer_count":82,"percentile":61.73,"percentile_exact":{"denominator":81,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","included":true,"model_id":"m-db208dec31031420","original_version_label":"4.0","peer_count":76,"percentile":54.67,"percentile_exact":{"denominator":3,"numerator":164},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","included":true,"model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":78,"percentile":58.44,"percentile_exact":{"denominator":77,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69}],"score":58.84,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":15950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1700},"family":"FrontierCode","row_ids":["ca9b3bebd5cf2b1f50a3f98180ff0076"],"score":34.69},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":200},"family":"Vals Code Migration","row_ids":["aeba1081b4abe0557cd65f41eca5f8b3"],"score":15.38}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","included":true,"model_id":"m-db208dec31031420","original_version_label":"2.0","peer_count":50,"percentile":34.69,"percentile_exact":{"denominator":49,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","included":true,"model_id":"m-db208dec31031420","original_version_label":"1.1","peer_count":27,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89}],"score":25.04,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":136952,"numerator":8496475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1375},"family":"GDP.pdf","row_ids":["7048e36b711cfdf67298d82bfa3d5f86"],"score":72.37},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1100},"family":"SimpleQA Verified","row_ids":["2fc7380a8d1e0cbfdc6837142943acdd"],"score":64.71},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"MultiChallenge","row_ids":["fc1a81b2311badda3a38cf9a6a28b5f4"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3900},"family":"AA-Omniscience","row_ids":["dbcce4096cefcfd79cea60ad99bf1516","0b6767461c06c4b95e61401653221d25"],"score":73.58}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","included":true,"model_id":"m-db208dec31031420","original_version_label":"current AA protocol","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","included":true,"model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","included":true,"model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412}],"score":62.04,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":24700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1000},"family":"Humanity Sixth Sense","row_ids":["fe8e2c3470907e5adda03a06d7b5f363"],"score":37.04},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1200},"family":"MMMU-Pro","row_ids":["71c04ca8ce2b90585afee7d563834b34"],"score":70.59}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","included":true,"model_id":"m-db208dec31031420","original_version_label":"current public protocol","peer_count":28,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","included":true,"model_id":"m-db208dec31031420","original_version_label":"current Standard 10","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295}],"score":53.81,"weight":10}],"eligible":true,"id":"m-db208dec31031420","in_window":true,"name":"Muse Spark Open","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":47,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"reasoning","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"c2edb36eb77a36c2b7e3b5065aafdd4c","included":true,"model_id":"m-db208dec31031420","original_version_label":"Tier 4 v2 / run 2.1.0","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"epoch-frontiermath-reviewed-method-20261029-task-version-2.1.0","protocol_id":"df0b5ec86237f8f4a709ac95","raw_value":0.3170731707317073,"reference_revision":"9a3445dafdfb59a2a1000d4a511e064525c84de4c1b7a45166c8fccae7b38fd7","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"43ea51951f3cc86c6ee38c3d"},"source_name":"Epoch AI","source_protocol_key":"epoch-frontiermath-reviewed-method-task-version-2.1.0","source_record_id":"43ea51951f3cc86c6ee38c3d","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.3170731707317073},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bf4b9ff10398e116bbd0ed6eb4a0a88d","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":67.8,"percentile_exact":{"denominator":59,"numerator":4000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.88989898989899,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-spark-open:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.88989898989899},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"356d1ac12dc91c2d1cadb615b7e3023b","included":true,"model_id":"m-db208dec31031420","original_version_label":"0.1","peer_count":53,"percentile":46.15,"percentile_exact":{"denominator":13,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.019047619047619,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-spark-open:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.019047619047619},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"muse-spark-open","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"04518aae51e855cbdd25679b26d57f51","included":true,"model_id":"m-db208dec31031420","original_version_label":"Diamond","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":28,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"36e4c3aab2f83b1d2c7a83a2"},"source_name":"Scale AI","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"36e4c3aab2f83b1d2c7a83a2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":28},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3eef34fe8a68f3f65cb23f07be80c5e4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-spark-open:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c140b9f2f3b8b047f022d22a6670f4c4","included":true,"model_id":"m-db208dec31031420","original_version_label":"v2.1","peer_count":82,"percentile":61.73,"percentile_exact":{"denominator":81,"numerator":5000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1401.12,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-spark-open:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1401.12},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d1fe83c673cc3e53108a1f798f9857e","included":true,"model_id":"m-db208dec31031420","original_version_label":"4.0","peer_count":76,"percentile":54.67,"percentile_exact":{"denominator":3,"numerator":164},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0808080808080808,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-spark-open:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0808080808080808},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cc297748bec17bb0b4aa16e861980849","included":true,"model_id":"m-db208dec31031420","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4970811827313194,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-spark-open:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4970811827313194},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"00f62480bdc3ab7dea00bbda96ff9331","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":78,"percentile":58.44,"percentile_exact":{"denominator":77,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1278.69,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-spark-open:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1278.69},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"ca9b3bebd5cf2b1f50a3f98180ff0076","included":true,"model_id":"m-db208dec31031420","original_version_label":"2.0","peer_count":50,"percentile":34.69,"percentile_exact":{"denominator":49,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0760421717171717,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-spark-open:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0760421717171717},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Muse Spark Open / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"aeba1081b4abe0557cd65f41eca5f8b3","included":true,"model_id":"m-db208dec31031420","original_version_label":"1.1","peer_count":27,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":26.89,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"26.89%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row24","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":26.89},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7048e36b711cfdf67298d82bfa3d5f86","included":true,"model_id":"m-db208dec31031420","original_version_label":"current AA protocol","peer_count":77,"percentile":72.37,"percentile_exact":{"denominator":19,"numerator":1375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.186,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-spark-open:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.186},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Muse-Spark-Open-816B-A42B_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"2fc7380a8d1e0cbfdc6837142943acdd","included":true,"model_id":"m-db208dec31031420","original_version_label":"run 1.2.0","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.566,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-15","result_timestamp":"2026-10-15T18:52:04.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"62f5f135855372af170f197a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"62f5f135855372af170f197a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.566},{"benchmark_family":"MultiChallenge","benchmark_id":"b-f4cf573847a266089c06","benchmark_label":"MultiChallenge","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:28:22.342721+00:00","component_key":"score","configuration":"Muse Spark Open","dimension":"language_information","direction":"higher_better","feed_id":"public-scale-multichallenge","id":"fc1a81b2311badda3a38cf9a6a28b5f4","included":true,"model_id":"m-db208dec31031420","original_version_label":"current LLM protocol","peer_count":5,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"scale-multichallenge-reviewed-method-20261029","protocol_id":"668b5eec1bd41046308d40fb","raw_value":74.2,"reference_revision":"5426e127548f994463014a8ce8eff3b89cdaaa891ad2a429a92160f494735493","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"6f318601d8452e4580471e20103281bb902fc361b85f144efabad1c260efa5a7","result_date":null,"source_cell":null,"source_id":"s-1643719e2a33872a6932","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"cf86936e5a687da2c2c01ad7"},"source_name":"Scale AI","source_protocol_key":"scale-multichallenge-reviewed-method","source_record_id":"cf86936e5a687da2c2c01ad7","source_url":"https://labs.scale.com/leaderboard/multichallenge?tab=large","unit":"%","value":74.2},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"dbcce4096cefcfd79cea60ad99bf1516","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":71.7,"percentile_exact":{"denominator":53,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.42033333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.42033333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Spark Open (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0b6767461c06c4b95e61401653221d25","included":true,"model_id":"m-db208dec31031420","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7199539965497412,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-spark-open:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7199539965497412},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Muse Spark Open (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"fe8e2c3470907e5adda03a06d7b5f363","included":true,"model_id":"m-db208dec31031420","original_version_label":"current public protocol","peer_count":28,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":29.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8459234783d71cd122818bb8"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"8459234783d71cd122818bb8","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":29.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Spark Open (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"71c04ca8ce2b90585afee7d563834b34","included":true,"model_id":"m-db208dec31031420","original_version_label":"current Standard 10","peer_count":35,"percentile":70.59,"percentile_exact":{"denominator":17,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.809248554913295,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-spark-open:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-spark-open:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.809248554913295}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"result_updated_at":"2026-10-15","score":48.62,"score_exact":{"denominator":9172073422512,"numerator":445978209583391},"source_id":"meta/muse-spark-open"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":13923,"numerator":703600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":700},"family":"Terminal-Bench-Science","row_ids":["e9c81f27a54c74df81f07409ddb33f17"],"score":53.85},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":300},"family":"HLE","row_ids":["433cbb1c2fcf649c36b004a451515cf8"],"score":42.86},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2800},"family":"CritPt","row_ids":["737809004b185d439bc3bdb639d99a34"],"score":54.9}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","included":true,"model_id":"m-60cd64beda366768","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","included":true,"model_id":"m-60cd64beda366768","original_version_label":"Diamond","peer_count":64,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295}],"score":50.54,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":1896048,"numerator":133400075},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":6100},"family":"GDPval-AA","row_ids":["42544e43967725b8e648b9b8615120d6"],"score":75.31},{"benchmark_version":"4.0","exact":{"denominator":12,"numerator":575},"family":"Terminal-Bench","row_ids":["3fbf327bd9cbc7d6e3195a9d909b9231"],"score":47.92},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1550},"family":"AutomationBench","row_ids":["250157187c56354b4eed518f47426340"],"score":81.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":5900},"family":"AA-Briefcase","row_ids":["e5699451024628001ea794ad1191793b"],"score":76.62}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2.1","peer_count":82,"percentile":75.31,"percentile_exact":{"denominator":81,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","included":true,"model_id":"m-60cd64beda366768","original_version_label":"4.0","peer_count":49,"percentile":47.92,"percentile_exact":{"denominator":12,"numerator":575},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","included":true,"model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":78,"percentile":76.62,"percentile_exact":{"denominator":77,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83}],"score":70.36,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":27425},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":2900},"family":"FrontierCode","row_ids":["d9ee15576eab5d8aeca7fb1035d14154"],"score":59.18},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":350},"family":"Vals Code Migration","row_ids":["501724f84396d924c42fcaca484ef573"],"score":26.92}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","included":true,"model_id":"m-60cd64beda366768","original_version_label":"2.0","peer_count":50,"percentile":59.18,"percentile_exact":{"denominator":49,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","included":true,"model_id":"m-60cd64beda366768","original_version_label":"1.1","peer_count":27,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55}],"score":43.05,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1007,"numerator":36975},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":750},"family":"GDP.pdf","row_ids":["0f9316648f89e3d521fa219b6361fd6c"],"score":39.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["084eb35df8b2a6c0ecbd814a3709a769","b4e27389bf022ba2e7fea370cfc23e71"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253}],"score":36.72,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":459,"numerator":10000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":700},"family":"Humanity Sixth Sense","row_ids":["672a5ff05bade6e76dd2270c17366a35"],"score":25.93},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":300},"family":"MMMU-Pro","row_ids":["f2f6f962e776bf31934ea15f004e8009"],"score":17.65}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current public protocol","peer_count":28,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289}],"score":21.79,"weight":10}],"eligible":true,"id":"m-60cd64beda366768","in_window":true,"name":"MiMo-V2.6-Flash","official_evidence":{"artifact_sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","basis":"official_release_page","date":"2026-09-22","sha256":"cf0895228ba87201c1b32c959410a72b49229245c631ef10fade5ed2e3e95b6b","url":"https://mimo.mi.com/docs/en-US/updates/model"},"primary_profile_id":"native","profile_id":"native","rank":48,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e9c81f27a54c74df81f07409ddb33f17","included":true,"model_id":"m-60cd64beda366768","original_version_label":"0.1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0428571428571429,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"mimo-v2-6-flash:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0428571428571429},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"433cbb1c2fcf649c36b004a451515cf8","included":true,"model_id":"m-60cd64beda366768","original_version_label":"Diamond","peer_count":64,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.137,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mimo-v2-6-flash:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.137},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"737809004b185d439bc3bdb639d99a34","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2 Hard","peer_count":52,"percentile":54.9,"percentile_exact":{"denominator":51,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.295,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"mimo-v2-6-flash:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.295},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42544e43967725b8e648b9b8615120d6","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v2.1","peer_count":82,"percentile":75.31,"percentile_exact":{"denominator":81,"numerator":6100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1613.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mimo-v2-6-flash:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1613.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"3fbf327bd9cbc7d6e3195a9d909b9231","included":true,"model_id":"m-60cd64beda366768","original_version_label":"4.0","peer_count":49,"percentile":47.92,"percentile_exact":{"denominator":12,"numerator":575},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":24.24,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"24.24%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":24.24},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"250157187c56354b4eed518f47426340","included":true,"model_id":"m-60cd64beda366768","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":81.58,"percentile_exact":{"denominator":19,"numerator":1550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.6405168250285657,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mimo-v2-6-flash:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6405168250285657},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5699451024628001ea794ad1191793b","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":78,"percentile":76.62,"percentile_exact":{"denominator":77,"numerator":5900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1486.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mimo-v2-6-flash:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1486.83},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d9ee15576eab5d8aeca7fb1035d14154","included":true,"model_id":"m-60cd64beda366768","original_version_label":"2.0","peer_count":50,"percentile":59.18,"percentile_exact":{"denominator":49,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.216152272727273,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"mimo-v2-6-flash:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.216152272727273},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"501724f84396d924c42fcaca484ef573","included":true,"model_id":"m-60cd64beda366768","original_version_label":"1.1","peer_count":27,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":39.55,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"39.55%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row20"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row20","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":39.55},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0f9316648f89e3d521fa219b6361fd6c","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current AA protocol","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.09,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mimo-v2-6-flash:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"084eb35df8b2a6c0ecbd814a3709a769","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":37.74,"percentile_exact":{"denominator":53,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.272,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.272},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiMo-V2.6-Flash","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b4e27389bf022ba2e7fea370cfc23e71","included":true,"model_id":"m-60cd64beda366768","original_version_label":"v1.1","peer_count":54,"percentile":30.19,"percentile_exact":{"denominator":53,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.47756410256410253,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mimo-v2-6-flash:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.47756410256410253},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"MiMo-v2.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"672a5ff05bade6e76dd2270c17366a35","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current public protocol","peer_count":28,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.4,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9eb6bb965bc2cf5d1fb85e61"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"9eb6bb965bc2cf5d1fb85e61","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.4},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiMo-V2.6-Flash","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"f2f6f962e776bf31934ea15f004e8009","included":true,"model_id":"m-60cd64beda366768","original_version_label":"current Standard 10","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.730635838150289,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mimo-v2-6-flash:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mimo-v2-6-flash:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.730635838150289}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":48.36,"score_exact":{"denominator":621835486272,"numerator":30069425541005},"source_id":"xiaomi/mimo-v2.6-flash"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1239,"numerator":75700},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4400},"family":"GPQA Diamond","row_ids":["7516363fe155a8bbdfc789cee4300773"],"score":74.58},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1000},"family":"HLE","row_ids":["7c461dbccbd2115e299061bd3cfb0f10"],"score":47.62}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":74.58,"percentile_exact":{"denominator":59,"numerator":4400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","peer_count":64,"percentile":47.62,"percentile_exact":{"denominator":21,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148}],"score":61.1,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":440154,"numerator":17585753},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":575},"family":"tau3-Banking","row_ids":["1b7ece54bb0355e69302fd5373585994"],"score":44.23},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1000},"family":"GDPval-AA","row_ids":["ae60755488af790d34ab9ed3aa726374"],"score":37.04},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["3126ca26ac0a6c7dab9a15033082575d"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":850},"family":"AutomationBench","row_ids":["2f9ec0e2d5074f31fb106a7f0db810e6"],"score":44.74},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":300},"family":"AA-Briefcase","row_ids":["8c6b16e6477c30b02c0cf413d3d98aff"],"score":27.27},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":400},"family":"EnterpriseOps-Gym","row_ids":["1d959564327d6d5496bf9174d774a4c8"],"score":44.44}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","peer_count":82,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","peer_count":78,"percentile":27.27,"percentile_exact":{"denominator":11,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996}],"score":39.95,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":2,"exact":{"denominator":3876,"numerator":157525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":225},"family":"IFBench","row_ids":["15ec6c548623aedd9b87ffa73167943d"],"score":56.25},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["60b6aea7e19a90a921731bc7d92e251c"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":300},"family":"SimpleQA Verified","row_ids":["29624d33497b02671ed8ec8a79e7f2e8"],"score":17.65}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365}],"score":40.64,"weight":20}],"eligible":true,"id":"m-5ea285ac7903cbc7","in_window":true,"name":"Kimi K2.7 Code","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":49,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7516363fe155a8bbdfc789cee4300773","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":74.58,"percentile_exact":{"denominator":59,"numerator":4400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.895959595959596,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"kimi-k2-7-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.895959595959596},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c461dbccbd2115e299061bd3cfb0f10","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"Diamond","peer_count":64,"percentile":47.62,"percentile_exact":{"denominator":21,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.148,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"kimi-k2-7-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.148},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1b7ece54bb0355e69302fd5373585994","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.202061855670103,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"kimi-k2-7-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.202061855670103},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae60755488af790d34ab9ed3aa726374","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v2.1","peer_count":82,"percentile":37.04,"percentile_exact":{"denominator":27,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1046.01,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"kimi-k2-7-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1046.01},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3126ca26ac0a6c7dab9a15033082575d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"kimi-k2-7-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2f9ec0e2d5074f31fb106a7f0db810e6","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.24454304615216713,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"kimi-k2-7-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.24454304615216713},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8c6b16e6477c30b02c0cf413d3d98aff","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"v1.1","peer_count":78,"percentile":27.27,"percentile_exact":{"denominator":11,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":858.04,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"kimi-k2-7-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":858.04},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1d959564327d6d5496bf9174d774a4c8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA run","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40226797970754996,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"kimi-k2-7-code:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40226797970754996},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"15ec6c548623aedd9b87ffa73167943d","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current 294-item protocol","peer_count":17,"percentile":56.25,"percentile_exact":{"denominator":4,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.631292517006803,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"kimi-k2-7-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631292517006803},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Kimi K2.7 Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"60b6aea7e19a90a921731bc7d92e251c","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"kimi-k2-7-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"kimi-k2-7-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k2.7-code","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"29624d33497b02671ed8ec8a79e7f2e8","included":true,"model_id":"m-5ea285ac7903cbc7","original_version_label":"run 1.2.0","peer_count":35,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.365,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:03.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d800e70f7011f6086e100151"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"d800e70f7011f6086e100151","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.365}],"release_date":"2026-06-12","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":47.7,"score_exact":{"denominator":43264497276,"numerator":2063781354715},"source_id":"moonshotai/kimi-k2.7-code"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":32214,"numerator":1462025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3600},"family":"GPQA Diamond","row_ids":["90978c171b499bb2efc92c2da0d369b4"],"score":61.02},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["5952fe944fd804ef75ab5466dc485304"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":250},"family":"HLE","row_ids":["7efb7f319954edefb37b62ce3d3ff826"],"score":35.71}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":61.02,"percentile_exact":{"denominator":59,"numerator":3600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116}],"score":45.38,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":4621617,"numerator":208057310},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"APEX-Agents","row_ids":["b5b0e9a07c5a8ced9fdf4c9ac74df7fc"],"score":50.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":775},"family":"tau3-Banking","row_ids":["84d4042d1c82bfee0ac491e53b4fa917"],"score":59.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4400},"family":"GDPval-AA","row_ids":["5a309e2224372571fd72b2e3de2fe6e1"],"score":54.32},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["cafc26cd33f16c1828c33bad26ea21a7"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":925},"family":"AutomationBench","row_ids":["2d900bebd8be5f37326880467236361d"],"score":48.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3400},"family":"AA-Briefcase","row_ids":["8d8e39fb86f2781e20fa7a7f68dc2ca4"],"score":44.16}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":59.62,"percentile_exact":{"denominator":13,"numerator":775},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","peer_count":82,"percentile":54.32,"percentile_exact":{"denominator":81,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":48.68,"percentile_exact":{"denominator":19,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","peer_count":78,"percentile":44.16,"percentile_exact":{"denominator":77,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16}],"score":45.02,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":38,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["ec9b7d6c1e90c4addeb4e33b7ca0403e"],"score":48.03}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"score":48.03,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":950},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":950},"family":"MMMU-Pro","row_ids":["d16dcb5ce825969190ea4f39de9f84b5"],"score":55.88}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867}],"score":55.88,"weight":10}],"eligible":true,"id":"m-d031501e2d0117d8","in_window":true,"name":"Apodex 1.1","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-30","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":50,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"90978c171b499bb2efc92c2da0d369b4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":61.02,"percentile_exact":{"denominator":59,"numerator":3600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.863636363636364,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"apodex-1-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.863636363636364},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5952fe944fd804ef75ab5466dc485304","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7efb7f319954edefb37b62ce3d3ff826","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"apodex-1-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b5b0e9a07c5a8ced9fdf4c9ac74df7fc","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA run","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.311946902654867,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"apodex-1-1:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.311946902654867},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84d4042d1c82bfee0ac491e53b4fa917","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":59.62,"percentile_exact":{"denominator":13,"numerator":775},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.251546391752577,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"apodex-1-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251546391752577},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a309e2224372571fd72b2e3de2fe6e1","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v2.1","peer_count":82,"percentile":54.32,"percentile_exact":{"denominator":81,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1195.61,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"apodex-1-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1195.61},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cafc26cd33f16c1828c33bad26ea21a7","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"apodex-1-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d900bebd8be5f37326880467236361d","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":48.68,"percentile_exact":{"denominator":19,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.28197929106327146,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"apodex-1-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28197929106327146},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8d8e39fb86f2781e20fa7a7f68dc2ca4","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"v1.1","peer_count":78,"percentile":44.16,"percentile_exact":{"denominator":77,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1012.16,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"apodex-1-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1012.16},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ec9b7d6c1e90c4addeb4e33b7ca0403e","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"apodex-1-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Apodex 1.1","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d16dcb5ce825969190ea4f39de9f84b5","included":true,"model_id":"m-d031501e2d0117d8","original_version_label":"current Standard 10","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.791907514450867,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"apodex-1-1:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"apodex-1-1:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.791907514450867}],"release_date":"2026-08-30","release_status":"published","release_verified":true,"result_updated_at":null,"score":47.24,"score_exact":{"denominator":148335419232,"numerator":7007785076525},"source_id":"published/aa:apodex-1-1"},{"benchmark_count":19,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":663,"numerator":45650},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["6de4c10b838de9bb43446d62c1dedc6c"],"score":65.38},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":200},"family":"HLE","row_ids":["ca6b622f3e2324e23758731c3eedb79c"],"score":66.67},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":3800},"family":"CritPt","row_ids":["a69315d97d8db2867eda5787e5741424"],"score":74.51}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"0.1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"Diamond","peer_count":64,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","peer_count":52,"percentile":74.51,"percentile_exact":{"denominator":51,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425}],"score":68.85,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":32706828,"numerator":1477535075},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":5200},"family":"GDPval-AA","row_ids":["f4a874da20b414e345395af6876bfede"],"score":64.2},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":175},"family":"Terminal-Bench","row_ids":["a8bde16affe89631785f57ec46935865"],"score":29.17},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1225},"family":"AutomationBench","row_ids":["c468f61930c57074fdae0429fd64cde2"],"score":64.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4700},"family":"AA-Briefcase","row_ids":["27c5dd58443826f50a40adf5e5fbdf05"],"score":61.04},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"OSWorld","row_ids":["b4bc853a9a856396a7313b82125218ee"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":1200},"family":"AA-AnalystAgent","row_ids":["ebd2172910fbd66a3c4a6655aa5bdef4"],"score":52.17}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2.1","peer_count":82,"percentile":64.2,"percentile_exact":{"denominator":81,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","included":true,"model_id":"m-e5d076977e951607","original_version_label":"4.0","peer_count":49,"percentile":29.17,"percentile_exact":{"denominator":6,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","included":true,"model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":78,"percentile":61.04,"percentile_exact":{"denominator":77,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.1","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","peer_count":24,"percentile":52.17,"percentile_exact":{"denominator":23,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375}],"score":45.18,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":4,"exact":{"denominator":1274,"numerator":32475},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3300},"family":"FrontierCode","row_ids":["91532a2952c890cc776404a1ca00195c"],"score":67.35},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":450},"family":"Vals Code Migration","row_ids":["1045763a4cc14f13d71cbc23b34d8b21"],"score":34.62},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["12d12a117fed600435582eb0cdccab38"],"score":0.0},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":0},"family":"MirrorCode","row_ids":["9b168b30e93ee9c993180224bf1ba997"],"score":0.0}],"family_count":4,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.0","peer_count":50,"percentile":67.35,"percentile_exact":{"denominator":49,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":27,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666}],"score":25.49,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":34238,"numerator":1755175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":3225},"family":"GDP.pdf","row_ids":["ff1f819163bf03455ec9cbe6d10872e3"],"score":84.87},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":450},"family":"SimpleQA Verified","row_ids":["d183987f58d811d89f8c42a38fe26df4"],"score":26.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["efcb1058be3817d672ab8961f07089a0","7ec06e89d9178a8c5c27508ed35805db"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906}],"score":51.26,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":51,"numerator":1000},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Humanity Sixth Sense","row_ids":["c23775cd50aa5f64341fa34e8b1cf6d3"],"score":0.0},{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1000},"family":"MMMU-Pro","row_ids":["7443158d8b6f105fcb7949451afd84ee"],"score":58.82},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["4a70f456ec596dc9670786d2fcc37f1a"],"score":0.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current public protocol","peer_count":28,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","included":true,"model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1}],"score":19.61,"weight":10}],"eligible":true,"id":"m-e5d076977e951607","in_window":true,"name":"GPT-6 Luna","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-22","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":51,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6de4c10b838de9bb43446d62c1dedc6c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"0.1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.0857142857142857,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gpt-6-luna:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0857142857142857},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ca6b622f3e2324e23758731c3eedb79c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"Diamond","peer_count":64,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.222,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gpt-6-luna:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.222},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a69315d97d8db2867eda5787e5741424","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2 Hard","peer_count":52,"percentile":74.51,"percentile_exact":{"denominator":51,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.425,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gpt-6-luna:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.425},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f4a874da20b414e345395af6876bfede","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v2.1","peer_count":82,"percentile":64.2,"percentile_exact":{"denominator":81,"numerator":5200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1444.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gpt-6-luna:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1444.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a8bde16affe89631785f57ec46935865","included":true,"model_id":"m-e5d076977e951607","original_version_label":"4.0","peer_count":49,"percentile":29.17,"percentile_exact":{"denominator":6,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":13.64,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"13.64%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row38"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row38","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":13.64},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c468f61930c57074fdae0429fd64cde2","included":true,"model_id":"m-e5d076977e951607","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":64.47,"percentile_exact":{"denominator":19,"numerator":1225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5316814034483175,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gpt-6-luna:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5316814034483175},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27c5dd58443826f50a40adf5e5fbdf05","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":78,"percentile":61.04,"percentile_exact":{"denominator":77,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1337.06,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gpt-6-luna:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1337.06},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"agents","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"b4bc853a9a856396a7313b82125218ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.1","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":48.9,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"48.9% Offline subset","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row6:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-cdb1d51fae04:table0:row6:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":48.9},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ebd2172910fbd66a3c4a6655aa5bdef4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA pass^5","peer_count":24,"percentile":52.17,"percentile_exact":{"denominator":23,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.4375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"gpt-6-luna:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"91532a2952c890cc776404a1ca00195c","included":true,"model_id":"m-e5d076977e951607","original_version_label":"2.0","peer_count":50,"percentile":67.35,"percentile_exact":{"denominator":49,"numerator":3300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.251747727272727,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gpt-6-luna:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.251747727272727},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"GPT-6 Luna / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"1045763a4cc14f13d71cbc23b34d8b21","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":27,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":46.67,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"46.67%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row18"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row18","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":46.67},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT-6 Luna (max)","dimension":"coding","direction":"higher_better","id":"12d12a117fed600435582eb0cdccab38","included":true,"model_id":"m-e5d076977e951607","original_version_label":"1.1","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official release prose explicitly gives max effort score","protocol_id":"2c719e8ae7701b7ffdc132b5","raw_value":66.6,"reference_revision":"85c3dd7a36e3bf7d239524921639deed18ddf859b0e45b09187bfa7759265e95","research_artifact_sha256":"10dca18a4eac16179eeec8e857211adacce367eccc788fba73f395fdc29070f7","response_sha256":"7fad5c2d996705fb830ee6da4bdf30fc1cea244b85083062ae8f1297902f274f","result_date":"2026-10-26","source_cell":null,"source_id":"s-38e2a40ddff67e804e8c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"GPT-6 Luna:DeepSWE:release-prose"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:Official release prose explicitly gives max effort score","source_record_id":"GPT-6 Luna:DeepSWE:release-prose","source_url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","unit":"%","value":66.6},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"9b168b30e93ee9c993180224bf1ba997","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 0.1.200","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.26666666666666666,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:48:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c853730ff624c3d4e4abe594"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"c853730ff624c3d4e4abe594","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.26666666666666666},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff1f819163bf03455ec9cbe6d10872e3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current AA protocol","peer_count":77,"percentile":84.87,"percentile_exact":{"denominator":38,"numerator":3225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.228,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gpt-6-luna:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gpt-6-luna_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"d183987f58d811d89f8c42a38fe26df4","included":true,"model_id":"m-e5d076977e951607","original_version_label":"run 1.2.0","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.414,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-09-22","result_timestamp":"2026-09-22T22:14:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"c39446deefa993659d55207a"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"c39446deefa993659d55207a","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.414},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"efcb1058be3817d672ab8961f07089a0","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3963333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3963333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"GPT-6 Luna (Max)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7ec06e89d9178a8c5c27508ed35805db","included":true,"model_id":"m-e5d076977e951607","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.20651573716178906,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gpt-6-luna:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.20651573716178906},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"c23775cd50aa5f64341fa34e8b1cf6d3","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current public protocol","peer_count":28,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"3b5b215e020d63c9bda1766e"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"3b5b215e020d63c9bda1766e","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"GPT-6 Luna (Max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7443158d8b6f105fcb7949451afd84ee","included":true,"model_id":"m-e5d076977e951607","original_version_label":"current Standard 10","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.796531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gpt-6-luna:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gpt-6-luna:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.796531791907514},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:28.329858+00:00","component_key":"score","configuration":"GPT-6 Luna","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-cdb1d51fae043110","id":"4a70f456ec596dc9670786d2fcc37f1a","included":true,"model_id":"m-e5d076977e951607","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-cdb1d51fae04; row Visual reasoning Chartography; preserve page footnotes and task-specific settings","protocol_id":"8ee4ee2a2791528dd623c2bb","raw_value":29.1,"reference_revision":"84edf72daf0a150e265b5bc8d39e78555bfb5a9b993cd8510e4afed6c52ba71b","research_artifact_sha256":"c3d2c9c235a74f372b9bd95fa875dbefb7ab7f8bc7279df1916247fe2b88473a","response_sha256":"d9203ea8fb2bd20681ee3b0868a11769bc37844f23ad0189c23d2fccd83d675b","result_date":"2026-10-07","source_cell":"29.1% no tools","source_id":"s-829fcf3ef09cdd6c033e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-cdb1d51fae04:table0:row15:column3"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual reasoning Chartography","source_record_id":"official-cdb1d51fae04:table0:row15:column3","source_url":"https://www.anthropic.com/claude-haiku-5-5","unit":"%","value":29.1}],"release_date":"2026-09-22","release_status":"published","release_verified":true,"result_updated_at":"2026-10-26","score":45.82,"score_exact":{"denominator":10726662138192,"numerator":491483807341325},"source_id":"openai/gpt-6-luna"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":2457,"numerator":80425},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":575},"family":"Terminal-Bench-Science","row_ids":["63428ab4d6f5aa0cb6c23626f378d1a8"],"score":44.23},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1300},"family":"HLE","row_ids":["5c428eec7c6c7eddbfd00d536e4c24e0"],"score":20.63},{"benchmark_version":"2 Hard","exact":{"denominator":3,"numerator":100},"family":"CritPt","row_ids":["2147a64a171deda69e1dcacdea28ad28"],"score":33.33}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"0.1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","peer_count":64,"percentile":20.63,"percentile_exact":{"denominator":63,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","peer_count":52,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155}],"score":32.73,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":4389,"numerator":275636},"families":[{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":72},"family":"Terminal-Bench","row_ids":["37a1dedbc40bda682abb927c484863ce"],"score":72.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1200},"family":"AutomationBench","row_ids":["87eb4175855e40447de186dd2f3411a6"],"score":63.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4100},"family":"AA-Briefcase","row_ids":["24d71c9e7d3f6a3d2ddf4b46e341dafa"],"score":53.25}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"4.0","peer_count":76,"percentile":72.0,"percentile_exact":{"denominator":1,"numerator":72},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":78,"percentile":53.25,"percentile_exact":{"denominator":77,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84}],"score":62.8,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1800},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1800},"family":"FrontierCode","row_ids":["9302ab59d95f9c2a82c7d7c38865f23b"],"score":36.73}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"2.0","peer_count":50,"percentile":36.73,"percentile_exact":{"denominator":49,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202}],"score":36.73,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":195525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["632aef0c9b2c1ee60a8fee169a7ead99"],"score":48.03},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2600},"family":"AA-Omniscience","row_ids":["464c122b5f2d778a04114dc266e1a6b8","d5ec18e90fffd664c6089d00e270f823"],"score":49.06}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698}],"score":48.54,"weight":20}],"eligible":true,"id":"m-8555d29b166d94e4","in_window":true,"name":"Nanbeige4.3-33B","official_evidence":{"basis":"authoritative_explicit_release_record","date":"2026-10-25","sha256":"2958e6615c6a96841d742ea7cd9cd0950fbea0ad208439a1436c7772261bca66","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":52,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63428ab4d6f5aa0cb6c23626f378d1a8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"0.1","peer_count":53,"percentile":44.23,"percentile_exact":{"denominator":13,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00952380952380952,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00952380952380952},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c428eec7c6c7eddbfd00d536e4c24e0","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"Diamond","peer_count":64,"percentile":20.63,"percentile_exact":{"denominator":63,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.069,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.069},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2147a64a171deda69e1dcacdea28ad28","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v2 Hard","peer_count":52,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.155,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.155},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"37a1dedbc40bda682abb927c484863ce","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"4.0","peer_count":76,"percentile":72.0,"percentile_exact":{"denominator":1,"numerator":72},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.297979797979798,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.297979797979798},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87eb4175855e40447de186dd2f3411a6","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":63.16,"percentile_exact":{"denominator":19,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.5311537657440716,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5311537657440716},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24d71c9e7d3f6a3d2ddf4b46e341dafa","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":78,"percentile":53.25,"percentile_exact":{"denominator":77,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1161.84,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1161.84},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9302ab59d95f9c2a82c7d7c38865f23b","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"2.0","peer_count":50,"percentile":36.73,"percentile_exact":{"denominator":49,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.10717702020202,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.10717702020202},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"632aef0c9b2c1ee60a8fee169a7ead99","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"464c122b5f2d778a04114dc266e1a6b8","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":5.66,"percentile_exact":{"denominator":53,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11666666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11666666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nanbeige4.3-33B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d5ec18e90fffd664c6089d00e270f823","included":true,"model_id":"m-8555d29b166d94e4","original_version_label":"v1.1","peer_count":54,"percentile":92.45,"percentile_exact":{"denominator":53,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8181132075471698,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8181132075471698}],"release_date":"2026-10-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":45.49,"score_exact":{"denominator":1714619907,"numerator":77994001630},"source_id":"published/aa:nanbeige4-3-33b"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":730184,"numerator":43081725},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":5500},"family":"GPQA Diamond","row_ids":["67157f1183905451cf4dded1391fab8e"],"score":93.22},{"benchmark_version":"0.1.0","exact":{"denominator":26,"numerator":1025},"family":"Terminal-Bench-Science","row_ids":["1a6a33cd4cc454e60378e0a0694263f3"],"score":39.42},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":1100},"family":"HLE","row_ids":["4912807a823c5e919553ca48ade2b67a"],"score":52.38},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2600},"family":"CritPt","row_ids":["3d5749511482d0fbdaddfa896055ac6a"],"score":50.98}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":93.22,"percentile_exact":{"denominator":59,"numerator":5500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond","peer_count":64,"percentile":52.38,"percentile_exact":{"denominator":21,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","peer_count":52,"percentile":50.98,"percentile_exact":{"denominator":51,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28}],"score":59.0,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":992107116,"numerator":27805451125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":375},"family":"tau3-Banking","row_ids":["0494d5d3e701ac0729bd55524064b2fa"],"score":28.85},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4600},"family":"GDPval-AA","row_ids":["7ee504c210f7e12209fb9a5011615203"],"score":56.79},{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":25},"family":"Terminal-Bench","row_ids":["8f361f3f1d062dfe3e449864fa84aec9"],"score":6.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":775},"family":"AutomationBench","row_ids":["c964aa700a761259aa3ab0adef555cba"],"score":40.79},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3700},"family":"AA-Briefcase","row_ids":["703518382f29bb0a80ddba827b81e22d"],"score":48.05},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"EnterpriseOps-Gym","row_ids":["df3f56d8282f0ce2c7d297df530d227a"],"score":11.11},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":100},"family":"AA-AnalystAgent","row_ids":["8cef68049263078fc3646fc52bd5e75a"],"score":4.35}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","included":true,"model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":28.85,"percentile_exact":{"denominator":13,"numerator":375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2.1","peer_count":82,"percentile":56.79,"percentile_exact":{"denominator":81,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","included":true,"model_id":"m-da421acad97d607b","original_version_label":"4.0","peer_count":49,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","included":true,"model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":78,"percentile":48.05,"percentile_exact":{"denominator":77,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA run","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","peer_count":24,"percentile":4.35,"percentile_exact":{"denominator":23,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1}],"score":28.03,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":21950},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":3000},"family":"FrontierCode","row_ids":["3f103c270013e8e3a189c86aaaac75c3"],"score":61.22},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":100},"family":"Vals Code Migration","row_ids":["fbec8a823687d43757fc0efa8ad26423"],"score":7.69}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"2.0","peer_count":50,"percentile":61.22,"percentile_exact":{"denominator":49,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","included":true,"model_id":"m-da421acad97d607b","original_version_label":"1.1","peer_count":27,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9}],"score":34.46,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":3021,"numerator":181675},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"IFBench","row_ids":["89b2619aee2f39a13fb5f1677cbe7c28"],"score":100.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":775},"family":"GDP.pdf","row_ids":["7e0fa158109c3646bd13a80e33ec79ee"],"score":40.79},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2100},"family":"AA-Omniscience","row_ids":["253af28f0e697d86e6b8e62f78eae062","e14051d6dbdd6fd61ec4748398bf8936"],"score":39.62}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043}],"score":60.14,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":750},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":750},"family":"MMMU-Pro","row_ids":["9161d7aea787f48463e1dc1ca6952542"],"score":44.12}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977}],"score":44.12,"weight":10}],"eligible":true,"id":"m-da421acad97d607b","in_window":true,"name":"MiniMax-M3","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":53,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67157f1183905451cf4dded1391fab8e","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":93.22,"percentile_exact":{"denominator":59,"numerator":5500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.929292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minimax-m3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.929292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1a6a33cd4cc454e60378e0a0694263f3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"0.1","peer_count":53,"percentile":39.42,"percentile_exact":{"denominator":26,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0.00476190476190476,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"minimax-m3:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00476190476190476},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4912807a823c5e919553ca48ade2b67a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"Diamond","peer_count":64,"percentile":52.38,"percentile_exact":{"denominator":21,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.16,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"minimax-m3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3d5749511482d0fbdaddfa896055ac6a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2 Hard","peer_count":52,"percentile":50.98,"percentile_exact":{"denominator":51,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.28,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"minimax-m3:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0494d5d3e701ac0729bd55524064b2fa","included":true,"model_id":"m-da421acad97d607b","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":28.85,"percentile_exact":{"denominator":13,"numerator":375},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.152577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"minimax-m3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.152577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ee504c210f7e12209fb9a5011615203","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v2.1","peer_count":82,"percentile":56.79,"percentile_exact":{"denominator":81,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1249.95,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minimax-m3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1249.95},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"MiniMax-M3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"8f361f3f1d062dfe3e449864fa84aec9","included":true,"model_id":"m-da421acad97d607b","original_version_label":"4.0","peer_count":49,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.01,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"1.01%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row53"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row53","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":1.01},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c964aa700a761259aa3ab0adef555cba","included":true,"model_id":"m-da421acad97d607b","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.21251257600749407,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"minimax-m3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21251257600749407},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"703518382f29bb0a80ddba827b81e22d","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":78,"percentile":48.05,"percentile_exact":{"denominator":77,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1089.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"minimax-m3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1089.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df3f56d8282f0ce2c7d297df530d227a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA run","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3210981796478663,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"minimax-m3:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3210981796478663},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8cef68049263078fc3646fc52bd5e75a","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA pass^5","peer_count":24,"percentile":4.35,"percentile_exact":{"denominator":23,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"minimax-m3:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"3f103c270013e8e3a189c86aaaac75c3","included":true,"model_id":"m-da421acad97d607b","original_version_label":"2.0","peer_count":50,"percentile":61.22,"percentile_exact":{"denominator":49,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.219031313131313,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"minimax-m3:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.219031313131313},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"MiniMax-M3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"fbec8a823687d43757fc0efa8ad26423","included":true,"model_id":"m-da421acad97d607b","original_version_label":"1.1","peer_count":27,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":20.9,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"20.90%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row26"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row26","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":20.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"89b2619aee2f39a13fb5f1677cbe7c28","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current 294-item protocol","peer_count":17,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.828571428571429,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minimax-m3:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.828571428571429},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e0fa158109c3646bd13a80e33ec79ee","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current AA protocol","peer_count":77,"percentile":40.79,"percentile_exact":{"denominator":19,"numerator":775},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.098,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"minimax-m3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.098},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"253af28f0e697d86e6b8e62f78eae062","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":75.47,"percentile_exact":{"denominator":53,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.44866666666666666,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.44866666666666666},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"MiniMax-M3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e14051d6dbdd6fd61ec4748398bf8936","included":true,"model_id":"m-da421acad97d607b","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.08645707376058043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"minimax-m3:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08645707376058043},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniMax-M3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"9161d7aea787f48463e1dc1ca6952542","included":true,"model_id":"m-da421acad97d607b","original_version_label":"current Standard 10","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.785549132947977,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minimax-m3:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minimax-m3:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.785549132947977}],"release_date":"2026-06-01","release_status":"published","release_verified":true,"result_updated_at":null,"score":45.09,"score_exact":{"denominator":32455029033504,"numerator":1463328516402595},"source_id":"minimax/MiniMax-M3"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":2800},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2800},"family":"GPQA Diamond","row_ids":["8763ad1982a06ca4a70e322ffe1ddbd4"],"score":47.46}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":47.46,"percentile_exact":{"denominator":59,"numerator":2800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434}],"score":47.46,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":243243,"numerator":13098550},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":950},"family":"tau3-Banking","row_ids":["479061fd58fe09f756e592ab99fa1642"],"score":73.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3800},"family":"GDPval-AA","row_ids":["d1a5193baa7d56820cb5874aa60b7c4a"],"score":46.91},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3200},"family":"AA-Briefcase","row_ids":["56f3576f6b7e7bada3d4ea81a3e974d2"],"score":41.56}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","peer_count":82,"percentile":46.91,"percentile_exact":{"denominator":81,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","peer_count":78,"percentile":41.56,"percentile_exact":{"denominator":77,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75}],"score":53.85,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":500},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":500},"family":"GDP.pdf","row_ids":["b0db480f387627aef52b7f9b3d6804ac"],"score":26.32}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052}],"score":26.32,"weight":20}],"eligible":true,"id":"m-9e1e0f605d47cf02","in_window":true,"name":"Motif 3","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":54,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8763ad1982a06ca4a70e322ffe1ddbd4","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":47.46,"percentile_exact":{"denominator":59,"numerator":2800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.834343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"motif-3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.834343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"479061fd58fe09f756e592ab99fa1642","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":73.08,"percentile_exact":{"denominator":13,"numerator":950},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.352577319587629,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"motif-3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.352577319587629},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1a5193baa7d56820cb5874aa60b7c4a","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v2.1","peer_count":82,"percentile":46.91,"percentile_exact":{"denominator":81,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1127.34,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"motif-3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1127.34},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"56f3576f6b7e7bada3d4ea81a3e974d2","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"v1.1","peer_count":78,"percentile":41.56,"percentile_exact":{"denominator":77,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":991.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"motif-3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":991.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Motif 3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b0db480f387627aef52b7f9b3d6804ac","included":true,"model_id":"m-9e1e0f605d47cf02","original_version_label":"current AA protocol","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.052,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"motif-3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"motif-3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.052}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.7,"score_exact":{"denominator":1908727821,"numerator":83411342375},"source_id":"motif-technologies/motif-3"},{"benchmark_count":3,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":50.0,"evaluator_count":1,"exact":{"denominator":59,"numerator":2300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2300},"family":"GPQA Diamond","row_ids":["29813bdc52dc98722342b740c2f2d3f2"],"score":38.98}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":38.98,"percentile_exact":{"denominator":59,"numerator":2300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505}],"score":38.98,"weight":25},{"effective_weight_percent":50.0,"evaluator_count":1,"exact":{"denominator":234,"numerator":11275},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":675},"family":"tau3-Banking","row_ids":["76481bf229eeeddb8b44a15648197449"],"score":51.92},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":400},"family":"GDPval-AA","row_ids":["5d726bb21290d73d92a00bf9993c5a4a"],"score":44.44}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":51.92,"percentile_exact":{"denominator":13,"numerator":675},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","peer_count":82,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37}],"score":48.18,"weight":25}],"eligible":true,"id":"m-af0edba2f76c80d6","in_window":true,"name":"G9v3-39A5B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-20","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":55,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"29813bdc52dc98722342b740c2f2d3f2","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":38.98,"percentile_exact":{"denominator":59,"numerator":2300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.805050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"g9v3-39a5b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.805050505050505},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"76481bf229eeeddb8b44a15648197449","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":51.92,"percentile_exact":{"denominator":13,"numerator":675},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.220618556701031,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"g9v3-39a5b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.220618556701031},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"G9v3-39A5B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5d726bb21290d73d92a00bf9993c5a4a","included":true,"model_id":"m-af0edba2f76c80d6","original_version_label":"v2.1","peer_count":82,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1109.37,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"g9v3-39a5b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"g9v3-39a5b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1109.37}],"release_date":"2026-08-20","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.58,"score_exact":{"denominator":27612,"numerator":1203425},"source_id":"published/aa:g9v3-39a5b"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":38.46,"evaluator_count":2,"exact":{"denominator":24,"numerator":1025},"families":[{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":75},"family":"Terminal-Bench","row_ids":["a49b610cdf02973ee5519a3d58738257"],"score":18.75},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":200},"family":"AA-Briefcase","row_ids":["99d362c58056009c0378c89776bd8bfa"],"score":66.67}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","peer_count":49,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071}],"score":42.71,"weight":25},{"effective_weight_percent":30.77,"evaluator_count":2,"exact":{"denominator":4,"numerator":175},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["f4858ed8f9e084fde242d77a18dddeed"],"score":62.5},{"benchmark_version":"run 0.1.200","exact":{"denominator":1,"numerator":25},"family":"MirrorCode","row_ids":["09300d4eee5c7d673f52949c9b3dcb98"],"score":25.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224}],"score":43.75,"weight":20},{"effective_weight_percent":30.77,"evaluator_count":1,"exact":{"denominator":17,"numerator":750},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":750},"family":"SimpleQA Verified","row_ids":["5fb5dec681fa13f63b1b4d2e3c76490c"],"score":44.12}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483}],"score":44.12,"weight":20}],"eligible":true,"id":"m-b0c41bb29e51489d","in_window":true,"name":"Grok 4.5","official_evidence":{"artifact_sha256":"605d18dafa601ac99d2ffe30c675010636f4c489483187f70c548f59080af7bf","basis":"official_release_page","date":"2026-07-16","sha256":"605d18dafa601ac99d2ffe30c675010636f4c489483187f70c548f59080af7bf","url":"https://x.ai/news/grok-4-5"},"primary_profile_id":"native","profile_id":"native","rank":56,"records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Grok 4.5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"a49b610cdf02973ee5519a3d58738257","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"4.0","peer_count":49,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.59,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.59%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row44"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row44","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.59},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Grok 4.5 (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"99d362c58056009c0378c89776bd8bfa","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"v1.1","peer_count":19,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4070707070707071,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1d676712f38ce9f38c964c27"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"1d676712f38ce9f38c964c27","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4070707070707071},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Grok 4.5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"f4858ed8f9e084fde242d77a18dddeed","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column6"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column6","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"MirrorCode","benchmark_id":"b-d865c3ae07002c880c25","benchmark_label":"MirrorCode","benchmark_version":"run 0.1.200","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"coding","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"09300d4eee5c7d673f52949c9b3dcb98","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 0.1.200","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","protocol_id":"f59ad8767b1d0f2829ac677e","raw_value":0.32222222222222224,"reference_revision":"ead447aa7a4d41565faaea9fcf51588a5478348ba5d9f286754b8d1dc912de66","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-10-21","result_timestamp":"2026-10-21T16:49:00.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"11e4b52afb30be962bcd2c97"},"source_name":"Epoch AI","source_protocol_key":"mirrorcode-ml-private-2L-30tasks-3repeats-10Btokens-7days-high-task-version-0.1.200","source_record_id":"11e4b52afb30be962bcd2c97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.32222222222222224},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"grok-4.5_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"5fb5dec681fa13f63b1b4d2e3c76490c","included":true,"model_id":"m-b0c41bb29e51489d","original_version_label":"run 1.2.0","peer_count":35,"percentile":44.12,"percentile_exact":{"denominator":17,"numerator":750},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.483,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:29:30.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9de2984a92f30943cb1daace"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9de2984a92f30943cb1daace","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.483}],"release_date":"2026-07-16","release_status":"published","release_verified":true,"result_updated_at":"2026-10-21","score":43.46,"score_exact":{"denominator":5304,"numerator":230525},"source_id":"xai/grok-4.5"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":3100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3100},"family":"GPQA Diamond","row_ids":["012394f027c8418fe233e5f26b558306"],"score":52.54}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":52.54,"percentile_exact":{"denominator":59,"numerator":3100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545}],"score":52.54,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":118503,"numerator":6339076},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4700},"family":"GDPval-AA","row_ids":["87c313476c9a420ad27b41db7d0250bf"],"score":58.02},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":154},"family":"Terminal-Bench","row_ids":["8ad3da1ee13871ee0748612b7b0f2202"],"score":51.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1050},"family":"AutomationBench","row_ids":["f3bd4620346190964b5a6a49765359b6"],"score":55.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3800},"family":"AA-Briefcase","row_ids":["6f61748806659fcf7e221477d42b6563"],"score":49.35}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","peer_count":82,"percentile":58.02,"percentile_exact":{"denominator":81,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","peer_count":78,"percentile":49.35,"percentile_exact":{"denominator":77,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44}],"score":53.49,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":725},"family":"GDP.pdf","row_ids":["c038eca51adef7673b4117a8c612883d"],"score":38.16}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086}],"score":38.16,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":100},"family":"MMMU-Pro","row_ids":["90c2787f5b2a0845d431bc6af09c8c25"],"score":5.88}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746}],"score":5.88,"weight":10}],"eligible":true,"id":"m-2d46f5cc91cc119d","in_window":true,"name":"JT-4.1 Flash 236B A21B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-09","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":57,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"012394f027c8418fe233e5f26b558306","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":52.54,"percentile_exact":{"denominator":59,"numerator":3100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.845454545454545,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.845454545454545},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87c313476c9a420ad27b41db7d0250bf","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v2.1","peer_count":82,"percentile":58.02,"percentile_exact":{"denominator":81,"numerator":4700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1335.19,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"jt-4-1-flash-236b-a21b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1335.19},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"8ad3da1ee13871ee0748612b7b0f2202","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"jt-4-1-flash-236b-a21b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3bd4620346190964b5a6a49765359b6","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":55.26,"percentile_exact":{"denominator":19,"numerator":1050},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4647120946439391,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"jt-4-1-flash-236b-a21b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4647120946439391},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6f61748806659fcf7e221477d42b6563","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"v1.1","peer_count":78,"percentile":49.35,"percentile_exact":{"denominator":77,"numerator":3800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1092.44,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"jt-4-1-flash-236b-a21b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1092.44},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c038eca51adef7673b4117a8c612883d","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current AA protocol","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.086,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"jt-4-1-flash-236b-a21b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.086},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"JT-4.1 Flash 236B A21B (Non-reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"90c2787f5b2a0845d431bc6af09c8c25","included":true,"model_id":"m-2d46f5cc91cc119d","original_version_label":"current Standard 10","peer_count":35,"percentile":5.88,"percentile_exact":{"denominator":17,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.641040462427746,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"jt-4-1-flash-236b-a21b-non-reasoning:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.641040462427746}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.41,"score_exact":{"denominator":475434036,"numerator":20638975985},"source_id":"published/aa:jt-4-1-flash-236b-a21b"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":115841125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3900},"family":"GPQA Diamond","row_ids":["b588de5f182f53797aeaba57de63e608"],"score":66.1},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["2f6e14dbb9becc8bf0feb78f2973083f"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1900},"family":"HLE","row_ids":["03d73659cb7b3abbbd4921c8546a8418"],"score":30.16},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1400},"family":"CritPt","row_ids":["5f0a8e1454688084cb0e096162e1eff2"],"score":27.45}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":66.1,"percentile_exact":{"denominator":59,"numerator":3900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond","peer_count":64,"percentile":30.16,"percentile_exact":{"denominator":63,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","peer_count":52,"percentile":27.45,"percentile_exact":{"denominator":51,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12}],"score":35.25,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":446904872},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":900},"family":"tau3-Banking","row_ids":["348ef206dd0fa625708f257a200455f5"],"score":69.23},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4900},"family":"GDPval-AA","row_ids":["3637cd472b3e658153b3601c1c953f3e"],"score":60.49},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":48},"family":"Terminal-Bench","row_ids":["612528520707bc7ebd5f35810f7efeb6"],"score":48.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1000},"family":"AutomationBench","row_ids":["d30d4db6bfda99c8317f5eddf5d0e239"],"score":52.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4600},"family":"AA-Briefcase","row_ids":["4c2ed1cac020a530bd05f7d661a8c39c"],"score":59.74}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2.1","peer_count":82,"percentile":60.49,"percentile_exact":{"denominator":81,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"4.0","peer_count":76,"percentile":48.0,"percentile_exact":{"denominator":1,"numerator":48},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":78,"percentile":59.74,"percentile_exact":{"denominator":77,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48}],"score":58.02,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1900},"family":"FrontierCode","row_ids":["b9e8a0463a865f6221caaaf1d3aecc9d"],"score":38.78}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"2.0","peer_count":50,"percentile":38.78,"percentile_exact":{"denominator":49,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465}],"score":38.78,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1007,"numerator":40025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":650},"family":"GDP.pdf","row_ids":["65f6a64cea971ea25fa466fbee3340a6"],"score":34.21},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2400},"family":"AA-Omniscience","row_ids":["d414f5fe5dd8b0e7a11df833233b1edf","7d19cde6db2f8d3d5e0cb451c988a92a"],"score":45.28}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506}],"score":39.75,"weight":20}],"eligible":true,"id":"m-d2febb93118535a2","in_window":true,"name":"K2 Horizon 375B A23B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":58,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b588de5f182f53797aeaba57de63e608","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":66.1,"percentile_exact":{"denominator":59,"numerator":3900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.872727272727273,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-375b-a23b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.872727272727273},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f6e14dbb9becc8bf0feb78f2973083f","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"k2-horizon-375b-a23b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"03d73659cb7b3abbbd4921c8546a8418","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"Diamond","peer_count":64,"percentile":30.16,"percentile_exact":{"denominator":63,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.102,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"k2-horizon-375b-a23b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.102},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5f0a8e1454688084cb0e096162e1eff2","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2 Hard","peer_count":52,"percentile":27.45,"percentile_exact":{"denominator":51,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.12,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"k2-horizon-375b-a23b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.12},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"348ef206dd0fa625708f257a200455f5","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":69.23,"percentile_exact":{"denominator":13,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.342268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-375b-a23b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.342268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3637cd472b3e658153b3601c1c953f3e","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v2.1","peer_count":82,"percentile":60.49,"percentile_exact":{"denominator":81,"numerator":4900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1365.05,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-375b-a23b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1365.05},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612528520707bc7ebd5f35810f7efeb6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"4.0","peer_count":76,"percentile":48.0,"percentile_exact":{"denominator":1,"numerator":48},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0151515151515152,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-375b-a23b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0151515151515152},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d30d4db6bfda99c8317f5eddf5d0e239","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":52.63,"percentile_exact":{"denominator":19,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.3720802754388014,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-375b-a23b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3720802754388014},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c2ed1cac020a530bd05f7d661a8c39c","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":78,"percentile":59.74,"percentile_exact":{"denominator":77,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1292.48,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-375b-a23b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1292.48},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"b9e8a0463a865f6221caaaf1d3aecc9d","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"2.0","peer_count":50,"percentile":38.78,"percentile_exact":{"denominator":49,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.118471464646465,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"k2-horizon-375b-a23b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.118471464646465},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65f6a64cea971ea25fa466fbee3340a6","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"current AA protocol","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.074,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-375b-a23b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.074},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d414f5fe5dd8b0e7a11df833233b1edf","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18633333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18633333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K2 Horizon 375B A23B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7d19cde6db2f8d3d5e0cb451c988a92a","included":true,"model_id":"m-d2febb93118535a2","original_version_label":"v1.1","peer_count":54,"percentile":69.81,"percentile_exact":{"denominator":53,"numerator":3700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7095452683326506,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-375b-a23b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"k2-horizon-375b-a23b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7095452683326506}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":43.36,"score_exact":{"denominator":41274330401304,"numerator":1789607818257619},"source_id":"published/aa:k2-horizon-375b-a23b"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":342,"numerator":12275},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":250},"family":"GPQA Diamond","row_ids":["dd3025c1130213f154e327d038efe0ed"],"score":83.33},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["f2b3352d05d0abb45473dcad237d2a14"],"score":11.84},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":25},"family":"HLE","row_ids":["0e05cbe602caeb64b25d2230f75c2113"],"score":12.5}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","peer_count":17,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2}],"score":35.89,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":72,"numerator":3887},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":80},"family":"APEX-Agents","row_ids":["e4b1483bcb889a5ef567c4c02fc571dd"],"score":80.0},{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":275},"family":"Terminal-Bench","row_ids":["fff5c31ca1adab72967ff7c824d4653e"],"score":34.38},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":100},"family":"AutomationBench","row_ids":["cf571d5463813d8daec30e8a8f31740d"],"score":100.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":100},"family":"AA-Briefcase","row_ids":["04f715495ca6c55181e1b42a64312d78"],"score":11.11},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":400},"family":"MCP Atlas","row_ids":["b1436c94b33fa58416664a5d31b199a9"],"score":44.44}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3}],"score":53.99,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":13,"numerator":450},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":250},"family":"Vals Code Migration","row_ids":["549f9102d89b3e9a3355e2ef41ef0184"],"score":19.23},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"DeepSWE","row_ids":["c87699e8afdc2f67ab86923665dd9ed5"],"score":50.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","peer_count":27,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5}],"score":34.62,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":50},"family":"SimpleQA Verified","row_ids":["63b449d94d8f27d0e405ea5ec8384866"],"score":50.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506}],"score":50.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":3,"exact":{"denominator":81,"numerator":2600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":800},"family":"Humanity Sixth Sense","row_ids":["d100a26c5dab08a99d576ce25ec3994e"],"score":29.63},{"benchmark_version":"Standard 10","exact":{"denominator":3,"numerator":200},"family":"MMMU-Pro","row_ids":["9ea024ab5ca2b01fdc71c3f50f184c9c"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["24019ec9f1eb6cf5c33ae9562d1f1bac"],"score":0.0}],"family_count":3,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","peer_count":28,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1}],"score":32.1,"weight":10}],"eligible":true,"id":"m-9e460c49820cbe4e","in_window":true,"name":"Kimi K3","official_evidence":{"artifact_sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","basis":"official_release_page","date":"2026-07-16","sha256":"1db73c31aaef36052f569cee76707e86ec55f49ea863271df8bee7454f6f9de5","url":"https://www.kimi.com/en/blog/"},"primary_profile_id":"native","profile_id":"native","rank":59,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"dd3025c1130213f154e327d038efe0ed","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":83.33,"percentile_exact":{"denominator":3,"numerator":250},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":93.5,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"93.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":93.5},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"f2b3352d05d0abb45473dcad237d2a14","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row39"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row39","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:27:51.335281+00:00","component_key":"score","configuration":"kimi-k3","dimension":"reasoning","direction":"higher_better","feed_id":"public-scale-hle","id":"0e05cbe602caeb64b25d2230f75c2113","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Diamond","peer_count":17,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"scale-hle-reviewed-method-20261029","protocol_id":"de345739b732415f6fe3a72b","raw_value":22.2,"reference_revision":"06add1b381cafa5224ad567f5d66ea164ebe7033799e942f4c436edea9b7ae35","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"cf139055573f66e4b8b7a19ffb7840470add2464df916efc09211ad71bef1671","result_date":null,"source_cell":null,"source_id":"s-0a3b665f62276aebfb49","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"f9b6641eae4aaf3e8237e1d2"},"source_name":"Scale AI","source_notice":"Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.","source_protocol_key":"scale-hle-reviewed-method","source_record_id":"f9b6641eae4aaf3e8237e1d2","source_url":"https://labs.scale.com/leaderboard/hle-diamond","unit":"%","value":22.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"e4b1483bcb889a5ef567c4c02fc571dd","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Current published protocol","peer_count":6,"percentile":80.0,"percentile_exact":{"denominator":1,"numerator":80},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Kimi K3 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":41.0,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"41.0","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":41.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Kimi K3 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"fff5c31ca1adab72967ff7c824d4653e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"4.0","peer_count":49,"percentile":34.38,"percentile_exact":{"denominator":8,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":17.17,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"17.17%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row36"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row36","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":17.17},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"cf571d5463813d8daec30e8a8f31740d","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published developer protocol","peer_count":5,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":30.8,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"30.8","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":30.8},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Kimi K3 (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"04f715495ca6c55181e1b42a64312d78","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"v1.1","peer_count":19,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ab6ae9d5e1bcf70d95718c3f"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"ab6ae9d5e1bcf70d95718c3f","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"kimi-k3 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"b1436c94b33fa58416664a5d31b199a9","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.3,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e47073271a45448f3bead129"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"e47073271a45448f3bead129","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.3},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Kimi K3 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"549f9102d89b3e9a3355e2ef41ef0184","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1","peer_count":27,"percentile":19.23,"percentile_exact":{"denominator":13,"numerator":250},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":28.59,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"28.59%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row23"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row23","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":28.59},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"c87699e8afdc2f67ab86923665dd9ed5","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":67.5,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"67.5","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":67.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"kimi-k3_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"63b449d94d8f27d0e405ea5ec8384866","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"run 1.2.0","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.506,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:12.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"64ecba17813459365e457fa5"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"64ecba17813459365e457fa5","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.506},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Kimi-K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"d100a26c5dab08a99d576ce25ec3994e","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"current public protocol","peer_count":28,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":25.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"25e0fb3d0c06f1463bde823c"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"25e0fb3d0c06f1463bde823c","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":25.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Kimi K3 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9ea024ab5ca2b01fdc71c3f50f184c9c","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"Standard 10","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":81.6,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"81.6 / 83.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column1","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":81.6},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"K3","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"24019ec9f1eb6cf5c33ae9562d1f1bac","included":true,"model_id":"m-9e460c49820cbe4e","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row Chartography w/ tools (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"5aad08fc7a196bcc6b2b847f","raw_value":68.1,"reference_revision":"bf62ce7df2670be1321c791bf00e055406b18be4b8652d21a07a17d4dbbb5f6b","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"68.1","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row19:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Chartography:Chartography w/ tools (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row19:column3","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":68.1}],"release_date":"2026-07-16","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":42.6,"score_exact":{"denominator":640224,"numerator":27275101},"source_id":"moonshotai/kimi-k3"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":531,"numerator":21650},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3500},"family":"GPQA Diamond","row_ids":["c2e927593d58bea9e466eb1677b5bc62"],"score":59.32},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":200},"family":"HLE","row_ids":["4fb23f5b07fa697e50fa64d17b095add"],"score":22.22}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":59.32,"percentile_exact":{"denominator":59,"numerator":3500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond","peer_count":64,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076}],"score":40.77,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":64571609},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":925},"family":"tau3-Banking","row_ids":["a5800dd77226525d5b0005d3b9d5c59c"],"score":71.15},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4100},"family":"GDPval-AA","row_ids":["2d97f2a57324a678b4a67b9a04a099de"],"score":50.62},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["b740ee2b3c6f41e732abe95d92f63313"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":650},"family":"AutomationBench","row_ids":["1aac376daa38513de1520f91177fbd00"],"score":34.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3100},"family":"AA-Briefcase","row_ids":["3a94cd016eb68403ef51f517cb4b5722"],"score":40.26}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":71.15,"percentile_exact":{"denominator":13,"numerator":925},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v2.1","peer_count":82,"percentile":50.62,"percentile_exact":{"denominator":81,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v1.1","peer_count":78,"percentile":40.26,"percentile_exact":{"denominator":77,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95}],"score":41.91,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":700},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":700},"family":"GDP.pdf","row_ids":["8a1843eddbf9ae3836989a362277ba86"],"score":36.84}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084}],"score":36.84,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":800},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":800},"family":"MMMU-Pro","row_ids":["487c57fe9e6c668745daa6efed20c475"],"score":47.06}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543}],"score":47.06,"weight":10}],"eligible":true,"id":"m-724c2a784624afa2","in_window":true,"name":"Ling-3.0-flash-VL","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":60,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c2e927593d58bea9e466eb1677b5bc62","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":59.32,"percentile_exact":{"denominator":59,"numerator":3500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.861616161616162,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-flash-vl:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.861616161616162},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"4fb23f5b07fa697e50fa64d17b095add","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"Diamond","peer_count":64,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.076,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.076},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a5800dd77226525d5b0005d3b9d5c59c","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":71.15,"percentile_exact":{"denominator":13,"numerator":925},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.344329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-vl:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.344329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2d97f2a57324a678b4a67b9a04a099de","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v2.1","peer_count":82,"percentile":50.62,"percentile_exact":{"denominator":81,"numerator":4100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1169.42,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1169.42},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b740ee2b3c6f41e732abe95d92f63313","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1aac376daa38513de1520f91177fbd00","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":34.21,"percentile_exact":{"denominator":19,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1568437535870999,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1568437535870999},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3a94cd016eb68403ef51f517cb4b5722","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"v1.1","peer_count":78,"percentile":40.26,"percentile_exact":{"denominator":77,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":984.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":984.95},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8a1843eddbf9ae3836989a362277ba86","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current AA protocol","peer_count":77,"percentile":36.84,"percentile_exact":{"denominator":19,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.084,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.084},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-VL","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"487c57fe9e6c668745daa6efed20c475","included":true,"model_id":"m-724c2a784624afa2","original_version_label":"current Standard 10","peer_count":35,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.789595375722543,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-vl:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"ling-3-0-flash-vl:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.789595375722543}],"release_date":"2026-09-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":40.93,"score_exact":{"denominator":24722569872,"numerator":1011958813885},"source_id":"inclusionai/ling-3.0-flash-vl"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7,"numerator":400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":400},"family":"GPQA Diamond","row_ids":["11f1be50c8d04272ce6fde6b77e7d32d"],"score":57.14}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"reference_revision":"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3}],"score":57.14,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":3,"exact":{"denominator":21,"numerator":550},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"APEX-Agents","row_ids":["7155f49a20e936c90be455db70c3df42"],"score":33.33},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":50},"family":"Terminal-Bench","row_ids":["54a707812fad1f7aeac89c3fb45c1b29"],"score":16.67},{"benchmark_version":"current published task suite","exact":{"denominator":7,"numerator":200},"family":"AutomationBench","row_ids":["f68fab6bf158dd2600b8426c509f20a3"],"score":28.57}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"4.0","peer_count":49,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"reference_revision":"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1}],"score":26.19,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":2,"exact":{"denominator":91,"numerator":3350},"families":[{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":400},"family":"Vals Code Migration","row_ids":["8b32286773fba058c9bb87cdd6ac0c71"],"score":30.77},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":300},"family":"DeepSWE","row_ids":["49c0ea386ce0f4eee60fb157edec7d0d"],"score":42.86}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"1.1","peer_count":27,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3}],"score":36.81,"weight":20}],"eligible":true,"id":"m-383f40728e3e5259","in_window":true,"name":"Hy4 preview","official_evidence":{"artifact_sha256":"676214cf8f2ac19b63d1293254a5e1920f5c5b758546dd77d7bbb0d132589ea3","basis":"official_release_page","date":"2026-08-28","sha256":"676214cf8f2ac19b63d1293254a5e1920f5c5b758546dd77d7bbb0d132589ea3","url":"https://www.tencent.com/zh-hk/tencent-releases-and-open-sources-tencent-hy4-preview/"},"primary_profile_id":"native","profile_id":"native","rank":61,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"reasoning","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"11f1be50c8d04272ce6fde6b77e7d32d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":57.14,"percentile_exact":{"denominator":7,"numerator":400},"profile_key":"native","protocol_description":"GPQA Diamond; highest available reasoning; starred Tencent runs","protocol_id":"8ac3a853c44f6c551fe64a3e","raw_value":92.3,"reference_revision":"3255981dc47d993bc752af5a78b4148dd6dfe8e9880f2f7bb0f0158d3ccbf079","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"92.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"GPQA Diamond / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:GPQA Diamond","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":92.3},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents pass@1","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:04.884824+00:00","component_key":"score","configuration":"Hy4 preview / official APEX-Agents pass@1","dimension":"agents","direction":"higher_better","feed_id":"public-image-6cdc461567bf6d9d","id":"7155f49a20e936c90be455db70c3df42","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"Current published protocol","peer_count":7,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Original numeric labels visually read in official image; APEX-Agents; Official Hy4 preview blue series; not Hy3 cyan sub-series. JPEG read at decoder 1/8 scale, original SHA-256 preserved.","protocol_id":"9421676f6c30b2c29c7fe852","raw_value":37.1,"reference_revision":"5f14ee3447f287e67716a41ca7eec7a4e0d5db53237660fdbb24320df60db917","research_artifact_sha256":"6bfc6b454b479163ee1899b9d8e7e84197b82432673103c5b583c94cd8c2f77c","response_sha256":"1380ff67ebed3be637f0288ba7aa956ba4821ddc002100de44996271478c3fd4","result_date":null,"source_cell":null,"source_id":"s-ccebc4eeb29cd6842d0c","source_kind":"official_release_comparison","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark.jpg","record":"twelve-hy4-image:row6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents pass@1","source_record_id":"twelve-hy4-image:row6","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":37.1},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Hy4 preview / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"54a707812fad1f7aeac89c3fb45c1b29","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"4.0","peer_count":49,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":8.08,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"8.08%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row45"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row45","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":8.08},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"f68fab6bf158dd2600b8426c509f20a3","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"AutomationBench v1.0.6; highest available reasoning; starred Tencent runs","protocol_id":"0c0199cb037df16b91698054","raw_value":32.1,"reference_revision":"8d36f4e4be5035584931a91cbeba947c3667a688c30801fffa2f36441b7a54db","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"32.1","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"AutomationBench / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:AutomationBench","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":32.1},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Hy4 preview / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"8b32286773fba058c9bb87cdd6ac0c71","included":true,"model_id":"m-383f40728e3e5259","original_version_label":"1.1","peer_count":27,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":45.36,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"45.36%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row19","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":45.36},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy4 preview","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"49c0ea386ce0f4eee60fb157edec7d0d","included":true,"model_id":"m-383f40728e3e5259","peer_count":8,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":64.3,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"64.3","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy4 preview"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":64.3}],"release_date":"2026-08-28","release_status":"published","release_verified":true,"result_updated_at":null,"score":40.28,"score_exact":{"denominator":1911,"numerator":76975},"source_id":"tencent/hy4-preview"},{"benchmark_count":17,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":531,"numerator":32500},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4600},"family":"GPQA Diamond","row_ids":["b8284820623630cb7a5150e367b7bed4"],"score":77.97},{"benchmark_version":"Diamond","exact":{"denominator":9,"numerator":400},"family":"HLE","row_ids":["00361f2a88ddce87fbff29f4d32f2681"],"score":44.44}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":77.97,"percentile_exact":{"denominator":59,"numerator":4600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","peer_count":64,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139}],"score":61.21,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":13864851,"numerator":337939913},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["274b775cdefb96d48e1e4a5a4e4f52de"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["9202014a7bc9d24df89fe2534ffc35bc"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1900},"family":"GDPval-AA","row_ids":["c77b3a8d70f27e7507bc297339631f46"],"score":23.46},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["a84f629da93a24a910222138960c0c05"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":675},"family":"AutomationBench","row_ids":["1915dcb5be7d210d57854519bafaa4ef"],"score":35.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2600},"family":"AA-Briefcase","row_ids":["397f726491d251a65df1f043233be5de"],"score":33.77},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["e323a2908ddfad0f233fccef67ea0b29"],"score":50.0},{"benchmark_version":"4","exact":{"denominator":1,"numerator":0},"family":"BFCL","row_ids":["355d178390cc7668a17845fc6539c179"],"score":0.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["a05fb15039bcbd3145b6e77f10451f52"],"score":0.0}],"family_count":9,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","peer_count":82,"percentile":23.46,"percentile_exact":{"denominator":81,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","peer_count":78,"percentile":33.77,"percentile_exact":{"denominator":77,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2}],"score":24.37,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["9bf5f2dc87f09a1babb07763838a2e08"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":152,"numerator":9275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":175},"family":"IFBench","row_ids":["c4485d942b9b21c7a152b1888b11b7c4"],"score":87.5},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1075},"family":"GDP.pdf","row_ids":["850da6d9475999744f930f21d7861dd9"],"score":56.58},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"IFEval","row_ids":["5ec142fbe451bd07f915ecd120f2faec"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"MMLU-Pro","row_ids":["7be3425578af009ce953f32bc09ba9aa"],"score":0.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5}],"score":61.02,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1100},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":1100},"family":"MMMU-Pro","row_ids":["b6422caa44d095e13cde40b5c1a62bc3"],"score":64.71}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647}],"score":64.71,"weight":10}],"eligible":true,"id":"m-7190ddc26a5a37a0","in_window":true,"name":"Qwen3.7 Plus","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-01","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":62,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b8284820623630cb7a5150e367b7bed4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":77.97,"percentile_exact":{"denominator":59,"numerator":4600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.9,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"qwen3-7-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.9},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"00361f2a88ddce87fbff29f4d32f2681","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Diamond","peer_count":64,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.139,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"qwen3-7-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.139},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"274b775cdefb96d48e1e4a5a4e4f52de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.224188790560472,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"qwen3-7-plus:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.224188790560472},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9202014a7bc9d24df89fe2534ffc35bc","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"qwen3-7-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c77b3a8d70f27e7507bc297339631f46","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v2.1","peer_count":82,"percentile":23.46,"percentile_exact":{"denominator":81,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":774.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"qwen3-7-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":774.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a84f629da93a24a910222138960c0c05","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"qwen3-7-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1915dcb5be7d210d57854519bafaa4ef","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.17391220980015495,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"qwen3-7-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.17391220980015495},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"397f726491d251a65df1f043233be5de","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v1.1","peer_count":78,"percentile":33.77,"percentile_exact":{"denominator":77,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":915.35,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"qwen3-7-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":915.35},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e323a2908ddfad0f233fccef67ea0b29","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA run","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.40555058191584603,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"qwen3-7-plus:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.40555058191584603},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"355d178390cc7668a17845fc6539c179","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":72.9,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"72.9","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37-plus:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":72.9},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"agents","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"a05fb15039bcbd3145b6e77f10451f52","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":73.2,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"73.2","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37-plus:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":73.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"9bf5f2dc87f09a1babb07763838a2e08","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"1.1","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":16.5,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"16.5","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column3"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column3","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":16.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c4485d942b9b21c7a152b1888b11b7c4","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":87.5,"percentile_exact":{"denominator":2,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.779591836734694,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"qwen3-7-plus:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.779591836734694},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"850da6d9475999744f930f21d7861dd9","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current AA protocol","peer_count":77,"percentile":56.58,"percentile_exact":{"denominator":19,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.122,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"qwen3-7-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.122},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"5ec142fbe451bd07f915ecd120f2faec","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current original Google dataset","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.6,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"94.6","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37-plus:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":94.6},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:21.831981+00:00","component_key":"score","configuration":"Qwen3.7-Plus","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1809664d20064a1b","id":"7be3425578af009ce953f32bc09ba9aa","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Plus. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":88.5,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"070c06bd30f9ee764397980c4e5aac76538694817f96b7b809ed7bdb676bba94","response_sha256":"774da599eef4991ae68bd15e7d59b991eca39e92c74ca4918bac48da651d1aff","result_date":null,"source_cell":"88.5","source_id":"s-b389e587ad922a7e3323","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37-plus:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37-plus:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7-plus","unit":"%","value":88.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Qwen3.7 Plus","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b6422caa44d095e13cde40b5c1a62bc3","included":true,"model_id":"m-7190ddc26a5a37a0","original_version_label":"current Standard 10","peer_count":35,"percentile":64.71,"percentile_exact":{"denominator":17,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.804624277456647,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"qwen3-7-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"qwen3-7-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.804624277456647}],"release_date":"2026-06-01","release_status":"published","release_verified":true,"result_updated_at":"2026-08-26","score":40.07,"score_exact":{"denominator":111251564424,"numerator":4457774361283},"source_id":"alibaba/qwen3.7-plus"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":365092,"numerator":10439225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3000},"family":"GPQA Diamond","row_ids":["f63fe95853d236ed53d2b79d02b859cb"],"score":50.85},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["2c2a68b20b30379e3063400c424fa6d0"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":200},"family":"HLE","row_ids":["cda0d514d2e746d7e415df4602a30e61"],"score":28.57},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":300},"family":"CritPt","row_ids":["1b14f3e78b8a7ccb2a6b6d0eda26f9e2"],"score":17.65}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":50.85,"percentile_exact":{"denominator":59,"numerator":3000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","peer_count":64,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","peer_count":52,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035}],"score":28.59,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1027026,"numerator":39546907},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":450},"family":"tau3-Banking","row_ids":["62cc0b9af0e4241c44026a470cdd621c"],"score":34.62},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":100},"family":"GDPval-AA","row_ids":["9d93ac6bf71dd0301098e7d1b1db2fb3"],"score":33.33},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["54cd2ea0dc02b5047f97de0db07066fb"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":875},"family":"AutomationBench","row_ids":["bdc1d1c2838fa6785d600f455b24f788"],"score":46.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1500},"family":"AA-Briefcase","row_ids":["0ec16dc40db0414f2891438008ea0a1a"],"score":19.48},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":500},"family":"EnterpriseOps-Gym","row_ids":["5ffca0a973b63389e8c0ae797941349a"],"score":55.56}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","peer_count":82,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":46.05,"percentile_exact":{"denominator":19,"numerator":875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":78,"percentile":19.48,"percentile_exact":{"denominator":77,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188}],"score":38.51,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":1500},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1500},"family":"FrontierCode","row_ids":["26e16bd219d1e573463da6c55739a16d"],"score":30.61}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","peer_count":50,"percentile":30.61,"percentile_exact":{"denominator":49,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323}],"score":30.61,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2014,"numerator":119275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1175},"family":"GDP.pdf","row_ids":["b1d8307905906d9eadee1409ffb734e6"],"score":61.84},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":3000},"family":"AA-Omniscience","row_ids":["5ac3b36a409b29ee98389b739f2f25b7","4347afed89affe307401c55cb1f477b8"],"score":56.6}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043}],"score":59.22,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["ef305cf290c211e130604e205ee97f46"],"score":50.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624}],"score":50.0,"weight":10}],"eligible":true,"id":"m-e7f69d0dac9fe632","in_window":true,"name":"Gemini 3.5 Flash-Lite","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-21","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":63,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f63fe95853d236ed53d2b79d02b859cb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":50.85,"percentile_exact":{"denominator":59,"numerator":3000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.838383838383838,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemini-3-5-flash-lite:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.838383838383838},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2c2a68b20b30379e3063400c424fa6d0","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"gemini-3-5-flash-lite:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"cda0d514d2e746d7e415df4602a30e61","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"Diamond","peer_count":64,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.099,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"gemini-3-5-flash-lite:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.099},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b14f3e78b8a7ccb2a6b6d0eda26f9e2","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2 Hard","peer_count":52,"percentile":17.65,"percentile_exact":{"denominator":17,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.035,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"gemini-3-5-flash-lite:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.035},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"62cc0b9af0e4241c44026a470cdd621c","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":34.62,"percentile_exact":{"denominator":13,"numerator":450},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.175257731958763,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"gemini-3-5-flash-lite:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.175257731958763},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9d93ac6bf71dd0301098e7d1b1db2fb3","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v2.1","peer_count":82,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":994.52,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemini-3-5-flash-lite:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":994.52},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54cd2ea0dc02b5047f97de0db07066fb","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"gemini-3-5-flash-lite:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bdc1d1c2838fa6785d600f455b24f788","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":46.05,"percentile_exact":{"denominator":19,"numerator":875},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.25011270511840955,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"gemini-3-5-flash-lite:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.25011270511840955},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0ec16dc40db0414f2891438008ea0a1a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":78,"percentile":19.48,"percentile_exact":{"denominator":77,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"gemini-3-5-flash-lite:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.17},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5ffca0a973b63389e8c0ae797941349a","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA run","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.423455684870188,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"gemini-3-5-flash-lite:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.423455684870188},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"26e16bd219d1e573463da6c55739a16d","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"2.0","peer_count":50,"percentile":30.61,"percentile_exact":{"denominator":49,"numerator":1500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0602982323232323,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"gemini-3-5-flash-lite:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0602982323232323},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b1d8307905906d9eadee1409ffb734e6","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current AA protocol","peer_count":77,"percentile":61.84,"percentile_exact":{"denominator":19,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.136,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"gemini-3-5-flash-lite:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.136},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ac3b36a409b29ee98389b739f2f25b7","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":33.96,"percentile_exact":{"denominator":53,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.265,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Gemini 3.5 Flash-Lite","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4347afed89affe307401c55cb1f477b8","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"v1.1","peer_count":54,"percentile":79.25,"percentile_exact":{"denominator":53,"numerator":4200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7328798185941043,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"gemini-3-5-flash-lite:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7328798185941043},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemini 3.5 Flash-Lite","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ef305cf290c211e130604e205ee97f46","included":true,"model_id":"m-e7f69d0dac9fe632","original_version_label":"current Standard 10","peer_count":35,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.790173410404624,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemini-3-5-flash-lite:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemini-3-5-flash-lite:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.790173410404624}],"release_date":"2026-07-21","release_status":"published","release_verified":true,"result_updated_at":null,"score":39.74,"score_exact":{"denominator":3057357807504,"numerator":121505356698007},"source_id":"google/gemini-3.5-flash-lite"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":600},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["1632686674422c3a2065b7adb4e89430"],"score":31.58}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":31.58,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":200},"families":[{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":200},"family":"AA-Briefcase","row_ids":["ec2017cdf9726d6330e3b72963cbf545"],"score":22.22}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808}],"score":22.22,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":25},"family":"DeepSWE","row_ids":["2125ffdac9125e27a313037f91e5edb9"],"score":25.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0}],"score":25.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1325},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1325},"family":"SimpleQA Verified","row_ids":["8f6725165e10f2ce17308791e15f9d1c"],"score":77.94}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662}],"score":77.94,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":9,"numerator":500},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":500},"family":"Humanity Sixth Sense","row_ids":["6d9db035c8e86692a9eb736f0a2c02ca"],"score":55.56}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","peer_count":28,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9}],"score":55.56,"weight":10}],"eligible":true,"id":"m-77b5dbc317075e9d","in_window":true,"name":"Gemini 3.6 Flash","official_evidence":{"artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","basis":"official_release_page","date":"2026-07-21","sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/"},"primary_profile_id":"native","profile_id":"native","rank":64,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.6 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"1632686674422c3a2065b7adb4e89430","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.6 Flash (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"ec2017cdf9726d6330e3b72963cbf545","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"v1.1","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.2808080808080808,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ef412f1ff70349e09c08b53"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ef412f1ff70349e09c08b53","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.2808080808080808},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.6 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"2125ffdac9125e27a313037f91e5edb9","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"1.1","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":49.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"49%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column2"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column2","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":49.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.6-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"8f6725165e10f2ce17308791e15f9d1c","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0cf84adcc7addcdd126233d1"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0cf84adcc7addcdd126233d1","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.6-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6d9db035c8e86692a9eb736f0a2c02ca","included":true,"model_id":"m-77b5dbc317075e9d","original_version_label":"current public protocol","peer_count":28,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":31.9,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ec212007414ca4c114702f87"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"ec212007414ca4c114702f87","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":31.9}],"release_date":"2026-07-21","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":39.59,"score_exact":{"denominator":2907,"numerator":115100},"source_id":"google/gemini-3.6-flash"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":850},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["72ae1d1de2eb58dbc8e135c9056f08ba"],"score":44.74}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"score":44.74,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":4,"exact":{"denominator":96,"numerator":2275},"families":[{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":275},"family":"Terminal-Bench","row_ids":["349a8da908aa998fbc2c641cf58cfc1b"],"score":11.46},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["d209f552e4c2026bb776833fb9d113ca"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":50},"family":"AA-Briefcase","row_ids":["a4acfc86f781d098a642bc8bcfd57f42"],"score":16.67},{"benchmark_version":"1000 tasks","exact":{"denominator":3,"numerator":200},"family":"MCP Atlas","row_ids":["44434b3879169ad5c54a4316dc8731a9"],"score":66.67}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6}],"score":23.7,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["7b14fb9c23111ece0c4ecf6bf486054c"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":1325},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1325},"family":"SimpleQA Verified","row_ids":["f90868b4246f3217c1fa61fa9b95b789"],"score":77.94}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662}],"score":77.94,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":27,"numerator":1600},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":1600},"family":"Humanity Sixth Sense","row_ids":["1fedb85017624b14b5d56927cae80633"],"score":59.26}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","peer_count":28,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6}],"score":59.26,"weight":10}],"eligible":true,"id":"m-8d99798a7e24690c","in_window":true,"name":"Gemini 3.5 Flash","official_evidence":{"artifact_sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","basis":"official_release_page","date":"2026-05-19","sha256":"77681a995381741cd6d3a14a476379e9baa02c5247766642e325108d73efdcbf","url":"https://deepmind.google/models/model-cards/gemini-3-5-flash/"},"primary_profile_id":"native","profile_id":"native","rank":65,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"72ae1d1de2eb58dbc8e135c9056f08ba","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row22"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row22","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Gemini 3.5 Flash / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"349a8da908aa998fbc2c641cf58cfc1b","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"4.0","peer_count":49,"percentile":11.46,"percentile_exact":{"denominator":24,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":6.06,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"6.06%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row47"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row47","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":6.06},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"agents","direction":"higher_better","id":"d209f552e4c2026bb776833fb9d113ca","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"published developer protocol","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column10"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column10","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (Medium)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"a4acfc86f781d098a642bc8bcfd57f42","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"v1.1","peer_count":19,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.27635619242579323,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"77b1a3da37ec54ac1a1a1de6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"77b1a3da37ec54ac1a1a1de6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.27635619242579323},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Gemini 3.5 Flash (high)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"44434b3879169ad5c54a4316dc8731a9","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":83.6,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0f775ccf7875c42490326fc5"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"0f775ccf7875c42490326fc5","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":83.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Gemini 3.5 Flash","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"7b14fb9c23111ece0c4ecf6bf486054c","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":37.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"37%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column3"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column3","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":37.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"gemini-3.5-flash_high","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f90868b4246f3217c1fa61fa9b95b789","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"run 1.2.0","peer_count":35,"percentile":77.94,"percentile_exact":{"denominator":17,"numerator":1325},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.662,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:35.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"487a26546c147712ab96ec49"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"487a26546c147712ab96ec49","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.662},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Gemini-3.5-Flash (high)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"1fedb85017624b14b5d56927cae80633","included":true,"model_id":"m-8d99798a7e24690c","original_version_label":"current public protocol","peer_count":28,"percentile":59.26,"percentile_exact":{"denominator":27,"numerator":1600},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":32.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"75125a52b367d5a5f4b66752"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"75125a52b367d5a5f4b66752","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":32.6}],"release_date":"2026-05-19","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":38.62,"score_exact":{"denominator":1116288,"numerator":43114225},"source_id":"google/gemini-3.5-flash"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":61047,"numerator":2006075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["04366109cc25918e59cfe4140983a31b"],"score":11.84},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2500},"family":"HLE","row_ids":["8319710b72494f93151d3337f0dbcdfd"],"score":39.68},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":800},"family":"CritPt","row_ids":["bd76f518c9968d380eda9d90a689c8c2"],"score":47.06}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","peer_count":64,"percentile":39.68,"percentile_exact":{"denominator":63,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","peer_count":52,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265}],"score":32.86,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":158004,"numerator":7953793},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1400},"family":"GDPval-AA","row_ids":["522de4d527da35b85f7ead36d5bab4a3"],"score":51.85},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":154},"family":"Terminal-Bench","row_ids":["7a6b6215e85870e72209d73fd4101db4"],"score":51.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1125},"family":"AutomationBench","row_ids":["981d2354449aefd49dcd7a2e96840b24"],"score":59.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3000},"family":"AA-Briefcase","row_ids":["fb238aac3df6d482092b3981302c93d1"],"score":38.96}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","peer_count":82,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":59.21,"percentile_exact":{"denominator":19,"numerator":1125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":78,"percentile":38.96,"percentile_exact":{"denominator":77,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2}],"score":50.34,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":91,"numerator":1825},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":200},"family":"FrontierCode","row_ids":["06196596577df812dcdeed161680d16c"],"score":28.57},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":150},"family":"Vals Code Migration","row_ids":["baba3c8fb65a67f1bc8fc2ae1572a4d3"],"score":11.54}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","peer_count":50,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","peer_count":27,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85}],"score":20.05,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":98025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1025},"family":"GDP.pdf","row_ids":["25ac15cd539fc43d4a096d3dea9f7966"],"score":53.95},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["11dc653c29e2dbe67eaa7a540042fef3","111f3e2d8b49ec87b3bbe30f22c6238c"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239}],"score":48.67,"weight":20}],"eligible":true,"id":"m-b6a574cc5300d5f1","in_window":true,"name":"Beam 1.0","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-16","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":66,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Beam 1.0 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"04366109cc25918e59cfe4140983a31b","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"0.1.0","peer_count":39,"percentile":11.84,"percentile_exact":{"denominator":19,"numerator":225},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":1.43,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"1.43%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row37"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row37","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":1.43},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8319710b72494f93151d3337f0dbcdfd","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"Diamond","peer_count":64,"percentile":39.68,"percentile_exact":{"denominator":63,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.128,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"reflection-beam:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bd76f518c9968d380eda9d90a689c8c2","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2 Hard","peer_count":52,"percentile":47.06,"percentile_exact":{"denominator":17,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.265,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"reflection-beam:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.265},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"522de4d527da35b85f7ead36d5bab4a3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v2.1","peer_count":82,"percentile":51.85,"percentile_exact":{"denominator":27,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1177.08,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"reflection-beam:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1177.08},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7a6b6215e85870e72209d73fd4101db4","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"4.0","peer_count":76,"percentile":51.33,"percentile_exact":{"denominator":3,"numerator":154},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0303030303030303,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"reflection-beam:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0303030303030303},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"981d2354449aefd49dcd7a2e96840b24","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":59.21,"percentile_exact":{"denominator":19,"numerator":1125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.4845786522065588,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"reflection-beam:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4845786522065588},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fb238aac3df6d482092b3981302c93d1","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":78,"percentile":38.96,"percentile_exact":{"denominator":77,"numerator":3000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":976.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"reflection-beam:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":976.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"06196596577df812dcdeed161680d16c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"2.0","peer_count":50,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0553143939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"reflection-beam:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0553143939393939},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Beam 1.0 / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"baba3c8fb65a67f1bc8fc2ae1572a4d3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"1.1","peer_count":27,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":22.85,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"22.85%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row25"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row25","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":22.85},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"25ac15cd539fc43d4a096d3dea9f7966","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"current AA protocol","peer_count":77,"percentile":53.95,"percentile_exact":{"denominator":19,"numerator":1025},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.116,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"reflection-beam:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"11dc653c29e2dbe67eaa7a540042fef3","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":45.28,"percentile_exact":{"denominator":53,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.30733333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.30733333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Beam 1.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"111f3e2d8b49ec87b3bbe30f22c6238c","included":true,"model_id":"m-b6a574cc5300d5f1","original_version_label":"v1.1","peer_count":54,"percentile":41.51,"percentile_exact":{"denominator":53,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.5899903753609239,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"reflection-beam:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"reflection-beam:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.5899903753609239}],"release_date":"2026-10-16","release_status":"published","release_verified":true,"result_updated_at":null,"score":38.38,"score_exact":{"denominator":137088952,"numerator":5261997115},"source_id":"published/aa:reflection-beam"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":5525},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":50},"family":"FrontierMath","row_ids":["34d11ef45eddcdc63e92f205eb9eaa3c"],"score":50.0},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":75},"family":"GPQA Diamond","row_ids":["6b87188fa1e95a23261b2a5e267b3413"],"score":37.5},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":1100},"family":"Terminal-Bench-Science","row_ids":["9d8fe865edc8da70092f0b94926485a3"],"score":57.89}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0}],"score":48.46,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":12,"numerator":275},"families":[{"benchmark_version":"4.0","exact":{"denominator":4,"numerator":175},"family":"Terminal-Bench","row_ids":["7fbb0050443eb76fe54b1727ad2dc1fe"],"score":43.75},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"AutomationBench","row_ids":["e4ce479712f2cad43715215ba9603696"],"score":25.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"AA-Briefcase","row_ids":["56c8ab5d498b1025e9684eb22709b6c7"],"score":0.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","peer_count":49,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334}],"score":22.92,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":75},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":75},"family":"DeepSWE","row_ids":["99e7d048583a82e4719e8e299872ac27"],"score":37.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6}],"score":37.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":50},"family":"GDP.pdf","row_ids":["2f6ca6bd54419e351a2ba1a4c64dedd7"],"score":50.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7}],"score":50.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":200},"family":"Humanity Sixth Sense","row_ids":["2d9aead127b28a667d7f4b6c145f9945"],"score":7.41},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":50},"family":"MMMU-Pro","row_ids":["079f06724159c79071056bf02b730573"],"score":50.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","peer_count":28,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0}],"score":28.7,"weight":10}],"eligible":true,"id":"m-f8dbf8d16fd37aac","in_window":true,"name":"GPT-5.6 Terra","official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profile_id":"native","rank":67,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"34d11ef45eddcdc63e92f205eb9eaa3c","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v2","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":84.9,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":84.9},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"reasoning","direction":"higher_better","id":"6b87188fa1e95a23261b2a5e267b3413","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Terra. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.9,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.9%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"9d8fe865edc8da70092f0b94926485a3","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"0.1.0","peer_count":39,"percentile":57.89,"percentile_exact":{"denominator":19,"numerator":1100},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"10.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row19"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row19","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":10.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Terra / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"7fbb0050443eb76fe54b1727ad2dc1fe","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"4.0","peer_count":49,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":22.73,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"22.73%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":22.73},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"agents","direction":"higher_better","id":"e4ce479712f2cad43715215ba9603696","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"published developer protocol","peer_count":9,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.2,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Terra (Low)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"56c8ab5d498b1025e9684eb22709b6c7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"v1.1","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23333333333333334,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"78d9d382e93a00c331836bb6"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"78d9d382e93a00c331836bb6","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23333333333333334},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"coding","direction":"higher_better","id":"99e7d048583a82e4719e8e299872ac27","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"1.1","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":69.6,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":69.6},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"language_information","direction":"higher_better","id":"2f6ca6bd54419e351a2ba1a4c64dedd7","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current published protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":24.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":24.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Terra (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2d9aead127b28a667d7f4b6c145f9945","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"current public protocol","peer_count":28,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":22.7,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d76ecbd7b9558f9993f56b5"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"0d76ecbd7b9558f9993f56b5","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":22.7},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Terra","dimension":"multimodal_understanding","direction":"higher_better","id":"079f06724159c79071056bf02b730573","included":true,"model_id":"m-f8dbf8d16fd37aac","original_version_label":"Standard 10","peer_count":3,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (with tools); preserve page footnotes and task-specific settings","protocol_id":"f5c96106a212a39d97b2e63c","raw_value":82.0,"reference_revision":"b7e58f57d98e47694caf466beb4a30524b72714b31b3c307b20ded406d72cd82","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row2:column2"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (with tools)","source_record_id":"openai56:table6:row2:column2","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":82.0}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":"2026-07-09","score":38.22,"score_exact":{"denominator":8208,"numerator":313675},"source_id":"openai/gpt-5.6-terra"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1,"numerator":40},"families":[{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":40},"family":"GPQA Diamond","row_ids":["8afb3ab2b09d75bd6e3e04698e2c3cb8"],"score":40.0}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"reference_revision":"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4}],"score":40.0,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":18,"numerator":925},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["4ea5a3b77e03d03f1422a6728fc0742b"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":50},"family":"AA-Briefcase","row_ids":["680700051014bfbbbe521e958802f31c"],"score":5.56},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["0886e98ce0984c34f09c65e2184f8692"],"score":100.0},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":100},"family":"MCP Atlas","row_ids":["f29e77fd200946b25fe37a0df78120da"],"score":100.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"reference_revision":"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4}],"score":51.39,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["2206795ef677e458e6f834638793b28d"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"reference_revision":"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":3,"exact":{"denominator":34,"numerator":1945},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":70},"family":"IFBench","row_ids":["1be1bd504f0391002f7de4a46514d7c5"],"score":70.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"IFEval","row_ids":["98b4a1e95c1679056cdd34a216b7399d"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"MMLU-Pro","row_ids":["3e7111ab4e0b4463d98c531d78381779"],"score":100.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":1000},"family":"SimpleQA Verified","row_ids":["cb8036f627af27545731c0fdb0405c49"],"score":58.82}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","peer_count":6,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"score":57.21,"weight":20}],"eligible":true,"id":"m-a3a34c5608c68d32","in_window":true,"name":"Qwen3.7 Max","official_evidence":{"artifact_sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","basis":"official_release_page","date":"2026-05-21","sha256":"403e66a17ea6e8fac4c1bfeac8994d62e532487d4621088e5612533269955cc7","url":"https://www.alibabacloud.com/help/en/model-studio/newly-released-models"},"primary_profile_id":"native","profile_id":"native","rank":68,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"8afb3ab2b09d75bd6e3e04698e2c3cb8","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Diamond; fixed 198-question subset","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"8ef2772079cc631a5830d286","raw_value":92.4,"reference_revision":"24be0bc5791302966af5e62323deb2910c1d3dda8d7fef37504b8b8d2a46e314","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"92.4","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row25:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:GPQA Diamond:GPQA Diamond","source_record_id":"qwen38-max:table1:row25:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":92.4},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"4ea5a3b77e03d03f1422a6728fc0742b","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-max; row Automation-Bench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"8034c6f82ee403f4844044bd","raw_value":14.2,"reference_revision":"666f6bf0b74c9edcf97a30d28fb3aec45fdea5e8c7086a8d9f46269484ea20ad","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"14.2","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row20:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:AutomationBench:Automation-Bench (Pass@1)","source_record_id":"qwen38-max:table1:row20:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":14.2},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Qwen3.7 Max","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"680700051014bfbbbe521e958802f31c","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v1.1","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.23737373737373738,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"4bdc91178cef546e817705be"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"4bdc91178cef546e817705be","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.23737373737373738},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL-V4","benchmark_version":"4","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"0886e98ce0984c34f09c65e2184f8692","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL-V4; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a643d03be3b50ea260b8cf17","raw_value":75.0,"reference_revision":"bae84d422087cd613d6bc169935ffc9c6ed6e0b902b4e6610c44f099ed91d223","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"75.0","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row15:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:BFCL:BFCL-V4","source_record_id":"qwen37:table0:row15:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":75.0},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"agents","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"f29e77fd200946b25fe37a0df78120da","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"a64252d4a51c4ff3960d90c0","raw_value":76.4,"reference_revision":"4a0c005611dd86c3565f8d75158c146ceee6e274cad6e65592336ad9ab62c805","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"76.4","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row17:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MCP Atlas:MCP-Atlas","source_record_id":"qwen37:table0:row17:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":76.4},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"coding","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"2206795ef677e458e6f834638793b28d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"1.1","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page qwen38-weights; row DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"4c5c14ed1147651f8f95c7e1","raw_value":21.6,"reference_revision":"54bb22cede4550567c0d4b9c74a0fb461f4d92c4c0851a3b4708788f2555f8f9","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"21.6","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row4:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE 1.1","source_record_id":"qwen38-weights:table0:row4:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":21.6},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:25.534483+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-29578a85932db50f","id":"1be1bd504f0391002f7de4a46514d7c5","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"current published protocol","peer_count":6,"percentile":70.0,"percentile_exact":{"denominator":1,"numerator":70},"profile_key":"native","protocol_description":"Published page qwen38-max; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"af6f81a701fa5f1daec85fee","raw_value":79.1,"reference_revision":"2c723f8b9bec82a85542acd9519d8cc31ec948d0efe0f487179c070f31c1070c","research_artifact_sha256":"62ec0295413d0677d888329104007bcff17dac8755e843367585ea50a423d762","response_sha256":"5db4afe0a54bb5f8bd43fa529be0a8596a3c245cfbc36f7bb7aaaeac486c9059","result_date":null,"source_cell":"79.1","source_id":"s-ec2bc92acb8dcb2d7ded","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-max:table1:row27:column4"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFBench:IFBench","source_record_id":"qwen38-max:table1:row27:column4","source_url":"https://qwen.ai/blog?id=qwen3.8","unit":"%","value":79.1},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"98b4a1e95c1679056cdd34a216b7399d","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current original Google dataset","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"25e5cdaaeea049303cac90a0","raw_value":94.3,"reference_revision":"e2c9a6806e804195cc1c53ed28309c6836bbffb2674d4965aef1a5cb5efad5db","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"94.3","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row35:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:IFEval:IFEval","source_record_id":"qwen37:table0:row35:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":94.3},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:27:22.597752+00:00","component_key":"score","configuration":"Qwen3.7-Max","dimension":"language_information","direction":"higher_better","feed_id":"public-release-1c2f4e3b793a65d0","id":"3e7111ab4e0b4463d98c531d78381779","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column Qwen3.7-Max. Footnotes and all competing runs retained in source.","protocol_id":"def1c3d4f8bd615659844ee7","raw_value":89.6,"reference_revision":"f37abe0f36a7580a8750b596e4b6229d1dac8a56d3448f6540c71d1ef6384647","research_artifact_sha256":"7951cc8f22839de09c5ffdf6cd58026b6ecaf6c668d12685ba89b7b6dc70f893","response_sha256":"bbf5b0599254b96017da3b80e662fb0469787ce1935ea3dc8e292682d3fb9a3d","result_date":null,"source_cell":"89.6","source_id":"s-e0772012b51061bd757f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen37:table0:row32:column6"},"source_name":"qwen.ai","source_protocol_key":"official:qwen.ai:MMLU-Pro:MMLU-Pro","source_record_id":"qwen37:table0:row32:column6","source_url":"https://qwen.ai/blog?id=qwen3.7","unit":"%","value":89.6},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"qwen3.7-max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"cb8036f627af27545731c0fdb0405c49","included":true,"model_id":"m-a3a34c5608c68d32","original_version_label":"run 1.2.0","peer_count":35,"percentile":58.82,"percentile_exact":{"denominator":17,"numerator":1000},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.5576730190571715,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:34:09.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9cbcfec1ad0833361b94f889"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"9cbcfec1ad0833361b94f889","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.5576730190571715}],"release_date":"2026-05-21","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":38.1,"score_exact":{"denominator":5508,"numerator":209845},"source_id":"alibaba/qwen3.7-max"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":109575125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4100},"family":"GPQA Diamond","row_ids":["8e38fd38f9044aff514003f73b248716"],"score":69.49},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["a2c09e681871381cb89f80ae44384709"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":650},"family":"HLE","row_ids":["66e35510159ab65c6e04c83322148b65"],"score":10.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1850},"family":"CritPt","row_ids":["7de1f456199243fb6a03dcf75e8fd06c"],"score":36.27}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":69.49,"percentile_exact":{"denominator":59,"numerator":4100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18}],"score":33.35,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":58748329},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":725},"family":"tau3-Banking","row_ids":["87111691a6a525e89691d7ee783ff6e3"],"score":55.77},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3400},"family":"GDPval-AA","row_ids":["2235f722890b82a4475fc455a58c66ab"],"score":41.98},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["e5e132be2e915d3fe97356bbd06869f9"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":550},"family":"AutomationBench","row_ids":["ba2dbb617a983db0793fc0f1b006c6eb"],"score":28.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":3900},"family":"AA-Briefcase","row_ids":["1caa339c8531983b8c2298b6a2b6d27c"],"score":50.65}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":55.77,"percentile_exact":{"denominator":13,"numerator":725},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","peer_count":82,"percentile":41.98,"percentile_exact":{"denominator":81,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":28.95,"percentile_exact":{"denominator":19,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":78,"percentile":50.65,"percentile_exact":{"denominator":77,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32}],"score":38.13,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":2014,"numerator":86325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":625},"family":"GDP.pdf","row_ids":["75cb1248936655503fa568c6e5bc1e02"],"score":32.89},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2800},"family":"AA-Omniscience","row_ids":["fcae50a41df6fb08e7a7f301689104ab","9364a9aa01cea849d2eb414760bb16ec"],"score":52.83}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255}],"score":42.86,"weight":20}],"eligible":true,"id":"m-754df8b5922bf5da","in_window":true,"name":"Solar Pro 4","official_evidence":{"artifact_sha256":"4c3c4258c97035773c61772893086ae86059b2a773780452e3c61615afc1a64f","basis":"official_release_page","date":"2026-08-11","sha256":"4c3c4258c97035773c61772893086ae86059b2a773780452e3c61615afc1a64f","url":"https://www.upstage.ai/blog/en/solar-pro-4"},"primary_profile_id":"native","profile_id":"native","rank":69,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e38fd38f9044aff514003f73b248716","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":69.49,"percentile_exact":{"denominator":59,"numerator":4100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.890909090909091,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"solar-pro4:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.890909090909091},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a2c09e681871381cb89f80ae44384709","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"66e35510159ab65c6e04c83322148b65","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7de1f456199243fb6a03dcf75e8fd06c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87111691a6a525e89691d7ee783ff6e3","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":55.77,"percentile_exact":{"denominator":13,"numerator":725},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.232989690721649,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"solar-pro4:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.232989690721649},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2235f722890b82a4475fc455a58c66ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v2.1","peer_count":82,"percentile":41.98,"percentile_exact":{"denominator":81,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1105.73,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-pro4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1105.73},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e5e132be2e915d3fe97356bbd06869f9","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-pro4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ba2dbb617a983db0793fc0f1b006c6eb","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":28.95,"percentile_exact":{"denominator":19,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1359339457640078,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-pro4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1359339457640078},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1caa339c8531983b8c2298b6a2b6d27c","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":78,"percentile":50.65,"percentile_exact":{"denominator":77,"numerator":3900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1093.32,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-pro4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1093.32},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"75cb1248936655503fa568c6e5bc1e02","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"current AA protocol","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.07,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-pro4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-pro4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"fcae50a41df6fb08e7a7f301689104ab","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":18.87,"percentile_exact":{"denominator":53,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18433333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18433333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Pro 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"9364a9aa01cea849d2eb414760bb16ec","included":true,"model_id":"m-754df8b5922bf5da","original_version_label":"v1.1","peer_count":54,"percentile":86.79,"percentile_exact":{"denominator":53,"numerator":4600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7907642010625255,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7907642010625255}],"release_date":"2026-08-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":37.78,"score_exact":{"denominator":4586036711256,"numerator":173241935325145},"source_id":"upstage/solar-pro4"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":117650},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1300},"family":"GPQA Diamond","row_ids":["b5ca8a579b13fb2adc96c2c993971f7c"],"score":22.03},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2600},"family":"HLE","row_ids":["92b0eaaad67a910391a173c1c1816701"],"score":41.27}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":22.03,"percentile_exact":{"denominator":59,"numerator":1300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","peer_count":64,"percentile":41.27,"percentile_exact":{"denominator":63,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131}],"score":31.65,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":391169138},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":825},"family":"tau3-Banking","row_ids":["9ed3eb345da97e474521a1cd387dd7bc"],"score":63.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4300},"family":"GDPval-AA","row_ids":["95847d736743738d4e269c199a41aca9"],"score":53.09},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["84fdb274483242a5de60a1387cb766c1"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":825},"family":"AutomationBench","row_ids":["94674bbf056258753d15a53bade2e692"],"score":43.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":4000},"family":"AA-Briefcase","row_ids":["f0265d9e7d722340c4dcae3d534155b9"],"score":51.95}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":63.46,"percentile_exact":{"denominator":13,"numerator":825},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","peer_count":82,"percentile":53.09,"percentile_exact":{"denominator":81,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":43.42,"percentile_exact":{"denominator":19,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","peer_count":78,"percentile":51.95,"percentile_exact":{"denominator":77,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55}],"score":50.78,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1075},"family":"GDP.pdf","row_ids":["3fef1cbee3ec381cc5e40ecf79f229ab"],"score":28.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":28.29,"weight":20}],"eligible":true,"id":"m-71070d6294f8dc57","in_window":true,"name":"Quasar 438B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":70,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b5ca8a579b13fb2adc96c2c993971f7c","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":22.03,"percentile_exact":{"denominator":59,"numerator":1300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.732323232323232,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"quasar-438b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.732323232323232},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"92b0eaaad67a910391a173c1c1816701","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"Diamond","peer_count":64,"percentile":41.27,"percentile_exact":{"denominator":63,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.131,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"quasar-438b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9ed3eb345da97e474521a1cd387dd7bc","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":63.46,"percentile_exact":{"denominator":13,"numerator":825},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.28659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"quasar-438b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"95847d736743738d4e269c199a41aca9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v2.1","peer_count":82,"percentile":53.09,"percentile_exact":{"denominator":81,"numerator":4300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1189.7,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"quasar-438b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1189.7},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"84fdb274483242a5de60a1387cb766c1","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"quasar-438b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"94674bbf056258753d15a53bade2e692","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":43.42,"percentile_exact":{"denominator":19,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.23864439593997933,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"quasar-438b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23864439593997933},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f0265d9e7d722340c4dcae3d534155b9","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"v1.1","peer_count":78,"percentile":51.95,"percentile_exact":{"denominator":77,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1123.55,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"quasar-438b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1123.55},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Quasar 438B (Max, Based on GLM-5.2)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3fef1cbee3ec381cc5e40ecf79f229ab","included":true,"model_id":"m-71070d6294f8dc57","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"quasar-438b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"quasar-438b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"release_date":"2026-08-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":37.52,"score_exact":{"denominator":636242607,"numerator":23874290621},"source_id":"published/aa:quasar-438b"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["6d2464487e35f420dae74891c6e0dec0"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":513513,"numerator":11596703},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":400},"family":"tau3-Banking","row_ids":["f6ef999603d2cd2e7f3ecd5e4118aa3e"],"score":30.77},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":800},"family":"GDPval-AA","row_ids":["518222bbae6a6938c2fc68e86213e044"],"score":29.63},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["46a72c291abf86482bd30a1de9ec49ac"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":325},"family":"AutomationBench","row_ids":["3e7bd955b2670c77e55665a0183fa94c"],"score":17.11},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1700},"family":"AA-Briefcase","row_ids":["628eb864f68ddd810e173dc996c88c01"],"score":22.08}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","peer_count":82,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":17.11,"percentile_exact":{"denominator":19,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","peer_count":78,"percentile":22.08,"percentile_exact":{"denominator":77,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17}],"score":22.58,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1175},"family":"GDP.pdf","row_ids":["10af516d49f62278a451fea09c7878c1"],"score":30.92}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"score":30.92,"weight":20}],"eligible":true,"id":"m-d48cde87e1f9c28d","in_window":true,"name":"A.X-K2","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":71,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6d2464487e35f420dae74891c6e0dec0","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"a-x-k2:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f6ef999603d2cd2e7f3ecd5e4118aa3e","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":30.77,"percentile_exact":{"denominator":13,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.160824742268041,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"a-x-k2:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.160824742268041},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"518222bbae6a6938c2fc68e86213e044","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v2.1","peer_count":82,"percentile":29.63,"percentile_exact":{"denominator":27,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":968.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"a-x-k2:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":968.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"46a72c291abf86482bd30a1de9ec49ac","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"a-x-k2:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e7bd955b2670c77e55665a0183fa94c","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":17.11,"percentile_exact":{"denominator":19,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02479417400536003,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"a-x-k2:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02479417400536003},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"628eb864f68ddd810e173dc996c88c01","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"v1.1","peer_count":78,"percentile":22.08,"percentile_exact":{"denominator":77,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.17,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"a-x-k2:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.17},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"A.X-K2","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"10af516d49f62278a451fea09c7878c1","included":true,"model_id":"m-d48cde87e1f9c28d","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"a-x-k2:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"a-x-k2:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":36.88,"score_exact":{"denominator":424161738,"numerator":15641285435},"source_id":"published/aa:a-x-k2"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":57,"numerator":1375},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":50},"family":"GPQA Diamond","row_ids":["021a82f4f7be85e4b693852ce483ef03"],"score":16.67},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["5a0b6bca7e87cdf2b14ffe16a27cf187"],"score":31.58}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":24.12,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":3,"exact":{"denominator":54,"numerator":1975},"families":[{"benchmark_version":"4.0","exact":{"denominator":2,"numerator":75},"family":"Terminal-Bench","row_ids":["6bb0c331de392cf1dc35a1de5de30317"],"score":37.5},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["25c531e52c90edae0ad06f8953351733"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":650},"family":"AA-Briefcase","row_ids":["239fbb8019bce72c55d5b79e4b9186e5"],"score":72.22}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","peer_count":49,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516}],"score":36.57,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3,"numerator":200},"families":[{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":200},"family":"DeepSWE","row_ids":["72c84440fb70eb9d01c3241fc4d3e551"],"score":66.67}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4}],"score":66.67,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":51,"numerator":1150},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":100},"family":"IFBench","row_ids":["bb7586265eb9f4ee440c1de9eb72708e"],"score":33.33},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":200},"family":"SimpleQA Verified","row_ids":["f310a7a1d6e944a484a86a2598b55b31"],"score":11.76}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636}],"score":22.55,"weight":20}],"eligible":true,"id":"m-3535a3e7891af5ef","in_window":true,"name":"DeepSeek V4 Flash 0731","official_evidence":{"artifact_sha256":"3f80af72789dd0dd52b0bd5de471e2df2bc33c6eb443bf98eaeecc4dec6afbf4","basis":"official_release_page","date":"2026-07-31","sha256":"3f80af72789dd0dd52b0bd5de471e2df2bc33c6eb443bf98eaeecc4dec6afbf4","url":"https://api-docs.deepseek.com/updates/"},"primary_profile_id":"native","profile_id":"native","rank":72,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond (Pass@1)","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"021a82f4f7be85e4b693852ce483ef03","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond (Pass@1); exact column DS-V4-Flash. Footnotes and all competing runs retained in source.","protocol_id":"f442e5a9ad3210e681dbc34e","raw_value":89.9,"reference_revision":"e98ac1442f991ff0fd18f8c469eb9905c55051e3f03a0365f15845c0f74cffe3","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"89.9","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row2:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":89.9},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"5a0b6bca7e87cdf2b14ffe16a27cf187","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row28","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"6bb0c331de392cf1dc35a1de5de30317","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"4.0","peer_count":49,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":18.69,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"18.69%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row34"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row34","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":18.69},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:07.827026+00:00","component_key":"score","configuration":"DS-V4-Flash","dimension":"agents","direction":"higher_better","feed_id":"public-release-5c348986a15182f2","id":"25c531e52c90edae0ad06f8953351733","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"published developer protocol","peer_count":7,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page deepseek-v41-flash; row AutomationBench (Pass@1); preserve page footnotes and task-specific settings","protocol_id":"9312268968d4ddf4fd457042","raw_value":37.7,"reference_revision":"90f86a5b7a3dd75dbef080c57f7868c231a969556a3cab8385e3bdb8708c99cf","research_artifact_sha256":"d378f960af2ca4f2abc42da808ca02b620b1dc1afcb62446b2a618d6dc3b4a8e","response_sha256":"dee6adf859d5cdcdec19c57d9b321cad0d2c5b7ffb161aa3a613ea658c26b43e","result_date":null,"source_cell":"37.7","source_id":"s-ac219f0d5e0822daa3e4","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"deepseek-v41-flash:table1:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench (Pass@1)","source_record_id":"deepseek-v41-flash:table1:row17:column6","source_url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash","unit":"%","value":37.7},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"DeepSeek V4 Flash 0731 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"239fbb8019bce72c55d5b79e4b9186e5","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"v1.1","peer_count":19,"percentile":72.22,"percentile_exact":{"denominator":9,"numerator":650},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.41515151515151516,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"8ac586a602ba7230f6327081"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"8ac586a602ba7230f6327081","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.41515151515151516},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"coding","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"72c84440fb70eb9d01c3241fc4d3e551","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"1.1","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Agentic coding DeepSWE 1.1; preserve page footnotes and task-specific settings","protocol_id":"de8e60f94c1bc38e345e7379","raw_value":54.4,"reference_revision":"352d1c2b149b540afb3810b5e441d964a5f3135c36482458d96eb93a88395303","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"54.4","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row5:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:Agentic coding DeepSWE 1.1","source_record_id":"qwen38-flashnext:table0:row5:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":54.4},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:39.613876+00:00","component_key":"score","configuration":"DeepSeek-V4-Flash-0731","dimension":"language_information","direction":"higher_better","feed_id":"public-release-eb79656c925ccbff","id":"bb7586265eb9f4ee440c1de9eb72708e","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"current published protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Published page qwen38-flashnext; row Instruction following IFBench; preserve page footnotes and task-specific settings","protocol_id":"0711b41a3607466304f72d7f","raw_value":79.2,"reference_revision":"5590c5d2f0bfb90947496b11689c3cb360a310e915c97aa11f0aa525856c7fe1","research_artifact_sha256":"ec29fbabe49fd6db0704b535a6d92e7aa1ba2b5ea400008c76e6dccd4ff8953b","response_sha256":"d6cedb440688dbba3ca0d3c31287e5497c0ff82c2f383ad072ddeb4f0218f648","result_date":"2026-08-26","source_cell":"79.2","source_id":"s-4f49e51ca439fb40ae5c","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-flashnext:table0:row15:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:Instruction following IFBench","source_record_id":"qwen38-flashnext:table0:row15:column4","source_url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next","unit":"%","value":79.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"deepseek-v4-flash-0731_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"f310a7a1d6e944a484a86a2598b55b31","included":true,"model_id":"m-3535a3e7891af5ef","original_version_label":"run 1.2.0","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.33633633633633636,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:30:16.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"0d781974fafdd04dfabbb814"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"0d781974fafdd04dfabbb814","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.33633633633633636}],"release_date":"2026-07-31","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":36.69,"score_exact":{"denominator":313956,"numerator":11517775},"source_id":"deepseek/deepseek-v4-flash-0731"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":413,"numerator":23125},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4500},"family":"GPQA Diamond","row_ids":["9b6a2e4e945c90d8754fc862b2e6383b"],"score":76.27},{"benchmark_version":"Diamond","exact":{"denominator":7,"numerator":250},"family":"HLE","row_ids":["a0be97a286389603a66382ac4d794e81"],"score":35.71}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":76.27,"percentile_exact":{"denominator":59,"numerator":4500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116}],"score":55.99,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":1320462,"numerator":43715939},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"APEX-Agents","row_ids":["869a19a4a7bb34031859441309a0c7b5"],"score":0.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":700},"family":"tau3-Banking","row_ids":["a00dadd2de4cef1d6ffa67d4028686b0"],"score":53.85},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3100},"family":"GDPval-AA","row_ids":["3e2d18ecd2bf6e50f2e15788da94966d"],"score":38.27},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["87cb9fb97ee59be625938b752ce0bb36"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":725},"family":"AutomationBench","row_ids":["df8bf63c05795b45b4fd82cd57c8b504"],"score":38.16},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":400},"family":"AA-Briefcase","row_ids":["a684e0a09f0e331ba14d29191ea44f68"],"score":36.36}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","peer_count":82,"percentile":38.27,"percentile_exact":{"denominator":81,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","peer_count":78,"percentile":36.36,"percentile_exact":{"denominator":11,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24}],"score":33.11,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["6f7ac5ece00998e78042608639ac6e80"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["2081b40577b770d6210bb3099918429f"],"score":48.03}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"score":48.03,"weight":20}],"eligible":true,"id":"m-d23e83fbb1d3aabc","in_window":true,"name":"Hy3","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":73,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9b6a2e4e945c90d8754fc862b2e6383b","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":76.27,"percentile_exact":{"denominator":59,"numerator":4500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.896969696969697,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hy3:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.896969696969697},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a0be97a286389603a66382ac4d794e81","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"Diamond","peer_count":64,"percentile":35.71,"percentile_exact":{"denominator":7,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.116,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"hy3:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.116},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"869a19a4a7bb34031859441309a0c7b5","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"APEX-Agents pass@1; official Tracel agent; 250 steps; 4 CPUs; 16 GB RAM","protocol_id":"cc9a245346c3577b6adef0db","raw_value":24.4,"reference_revision":"e997e3956f9d2ce8564429b7beed9ce4290faaae3fa888842d874a68109a9d2d","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"24.4","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"APEX-Agents / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:APEX-Agents","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":24.4},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a00dadd2de4cef1d6ffa67d4028686b0","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":53.85,"percentile_exact":{"denominator":13,"numerator":700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.228865979381443,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hy3:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.228865979381443},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3e2d18ecd2bf6e50f2e15788da94966d","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v2.1","peer_count":82,"percentile":38.27,"percentile_exact":{"denominator":81,"numerator":3100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1070.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hy3:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1070.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"87cb9fb97ee59be625938b752ce0bb36","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hy3:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"df8bf63c05795b45b4fd82cd57c8b504","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":38.16,"percentile_exact":{"denominator":19,"numerator":725},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.18409128371172395,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"hy3:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18409128371172395},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a684e0a09f0e331ba14d29191ea44f68","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"v1.1","peer_count":78,"percentile":36.36,"percentile_exact":{"denominator":11,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":947.24,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hy3:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":947.24},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:34:18.208700+00:00","component_key":"score","configuration":"腾讯官方比较表 · Hy3","dimension":"coding","direction":"higher_better","feed_id":"public-tencent-hy4-appendix","id":"6f7ac5ece00998e78042608639ac6e80","included":true,"model_id":"m-d23e83fbb1d3aabc","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"DeepSWE current 1.1 task suite; mini-swe-agent; sandbox 8 CPUs and 16 GB RAM","protocol_id":"8ed6bc0f2ceb7f2b2ae817c0","raw_value":28.0,"reference_revision":"9a765ff00fcf8993e8f64b65ef9b85632c614cade299962a088e3173cea74910","response_sha256":"02c751bf51c52c52cc02094561d7144bbfe3e27084d48d8a405d580da3554e21","result_date":null,"source_cell":"28.0","source_id":"s-tencent-hy4-appendix-20261106","source_kind":"official_release","source_locator":{"image_url":"https://huggingface.co/tencent/Hy4-preview/resolve/main/assets/benchmark-appendix.jpg","record":"DeepSWE / Hy3"},"source_name":"腾讯 Hy4 官方发布比较表","source_protocol_key":"public-tencent-hy4-appendix:DeepSWE","source_url":"https://huggingface.co/tencent/Hy4-preview","unit":"%","value":28.0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Hy3","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2081b40577b770d6210bb3099918429f","included":true,"model_id":"m-d23e83fbb1d3aabc","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hy3:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"hy3:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112}],"release_date":"2026-07-06","release_status":"published","release_verified":true,"result_updated_at":null,"score":35.42,"score_exact":{"denominator":9816314508,"numerator":347716592885},"source_id":"tencent/hy3"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":1850},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1850},"family":"GPQA Diamond","row_ids":["6fa185e6f3a52003678d56e35ea14de0"],"score":31.36}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252}],"score":31.36,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":1539,"numerator":10700},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":700},"family":"GDPval-AA","row_ids":["60ed1cde923cee15c4935672b177aa62"],"score":8.64},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":100},"family":"AutomationBench","row_ids":["49cf1a20c5edfdf8e5e8bfd178db37ad"],"score":5.26}],"family_count":2,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","peer_count":82,"percentile":8.64,"percentile_exact":{"denominator":81,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258}],"score":6.95,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4,"numerator":325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":325},"family":"IFBench","row_ids":["af2b7a8fe076debb535a31ef7b09244c"],"score":81.25}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864}],"score":81.25,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":250},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":250},"family":"MMMU-Pro","row_ids":["ca0bc4ca214d2607b60ef7f92f73e72e"],"score":14.71}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514}],"score":14.71,"weight":10}],"eligible":true,"id":"m-a4be17f49ae915ce","in_window":true,"name":"Gemma 4 12B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":74,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"6fa185e6f3a52003678d56e35ea14de0","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"gemma-4-12b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"60ed1cde923cee15c4935672b177aa62","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"v2.1","peer_count":82,"percentile":8.64,"percentile_exact":{"denominator":81,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":434.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"gemma-4-12b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":434.79},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"49cf1a20c5edfdf8e5e8bfd178db37ad","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.006855723615934258,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006855723615934258},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af2b7a8fe076debb535a31ef7b09244c","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current 294-item protocol","peer_count":17,"percentile":81.25,"percentile_exact":{"denominator":4,"numerator":325},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.735374149659864,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"gemma-4-12b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.735374149659864},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Gemma 4 12B (Reasoning)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ca0bc4ca214d2607b60ef7f92f73e72e","included":true,"model_id":"m-a4be17f49ae915ce","original_version_label":"current Standard 10","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.696531791907514,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"gemma-4-12b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"gemma-4-12b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.696531791907514}],"release_date":"2026-06-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":34.12,"score_exact":{"denominator":24697872,"numerator":842744275},"source_id":"published/aa:gemma-4-12b"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":3300},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3300},"family":"GPQA Diamond","row_ids":["7c25d00a19511ccc369639cf5605291e"],"score":55.93}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657}],"score":55.93,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":183027373},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":500},"family":"tau3-Banking","row_ids":["3cd21f0f8001e029d7b19208e29ba099"],"score":38.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1700},"family":"GDPval-AA","row_ids":["2a88f9d5ec82d3f04004cf3da53914a5"],"score":20.99},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["7e0782bec5efa3fc907fbba656122a79"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":125},"family":"AutomationBench","row_ids":["bfaca7bcdefcec6c5d1daac93dc0d456"],"score":6.58},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1600},"family":"AA-Briefcase","row_ids":["d8f9d30bc51b5ac382358dbda4a533e5"],"score":20.78}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","peer_count":82,"percentile":20.99,"percentile_exact":{"denominator":81,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":78,"percentile":20.78,"percentile_exact":{"denominator":77,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95}],"score":23.76,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":12084,"numerator":238625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":75},"family":"IFBench","row_ids":["3080cdd2a4ddf6bfb1a2fed6805cd96e"],"score":18.75},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":375},"family":"GDP.pdf","row_ids":["5ecbc58b2efe8e4c7ad00b00fd409205"],"score":19.74},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1100},"family":"AA-Omniscience","row_ids":["f5ca62e67f09a75d5d460c81d0851fa4","2c99a44c194b274fedb14bdd3896ace9"],"score":20.75}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393}],"score":19.75,"weight":20}],"eligible":true,"id":"m-25ccaf97cf82a5e0","in_window":true,"name":"Ring-2.6-1T","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-08","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":75,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c25d00a19511ccc369639cf5605291e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":55.93,"percentile_exact":{"denominator":59,"numerator":3300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.856565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ring-2-6-1t:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.856565656565657},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3cd21f0f8001e029d7b19208e29ba099","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":38.46,"percentile_exact":{"denominator":13,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.179381443298969,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ring-2-6-1t:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.179381443298969},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2a88f9d5ec82d3f04004cf3da53914a5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v2.1","peer_count":82,"percentile":20.99,"percentile_exact":{"denominator":81,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":752.9,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ring-2-6-1t:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":752.9},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e0782bec5efa3fc907fbba656122a79","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ring-2-6-1t:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"bfaca7bcdefcec6c5d1daac93dc0d456","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.007178659181731186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ring-2-6-1t:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.007178659181731186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d8f9d30bc51b5ac382358dbda4a533e5","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":78,"percentile":20.78,"percentile_exact":{"denominator":77,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":657.95,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ring-2-6-1t:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":657.95},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3080cdd2a4ddf6bfb1a2fed6805cd96e","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current 294-item protocol","peer_count":17,"percentile":18.75,"percentile_exact":{"denominator":4,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.445578231292517,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"ring-2-6-1t:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.445578231292517},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5ecbc58b2efe8e4c7ad00b00fd409205","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"current AA protocol","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.038,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ring-2-6-1t:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f5ca62e67f09a75d5d460c81d0851fa4","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":32.08,"percentile_exact":{"denominator":53,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.263,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.263},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ring-2.6-1T","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2c99a44c194b274fedb14bdd3896ace9","included":true,"model_id":"m-25ccaf97cf82a5e0","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.154228855721393,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ring-2-6-1t:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ring-2-6-1t:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.154228855721393}],"release_date":"2026-05-08","release_status":"published","release_verified":true,"result_updated_at":null,"score":34.1,"score_exact":{"denominator":33720858171,"numerator":1150018852748},"source_id":"published/aa:ring-2-6-1t"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":7182,"numerator":275461},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1100},"family":"GDPval-AA","row_ids":["dde8616d382b2b0a447eb1104181da5e"],"score":40.74},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["7620c0fb57936004d0ff80edb6675f9f"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":800},"family":"AutomationBench","row_ids":["b1f8ac835c2c56f569ae7093d49e508f"],"score":42.11},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":200},"family":"AA-Briefcase","row_ids":["b76daa2b8d1a11ba3f7dd4198d33e0a0"],"score":28.57}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","peer_count":82,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":42.11,"percentile_exact":{"denominator":19,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":78,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74}],"score":38.35,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":4028,"numerator":113625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":925},"family":"GDP.pdf","row_ids":["d7e2c6c9bb5b766f6065fc5409408479"],"score":24.34},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1700},"family":"AA-Omniscience","row_ids":["8e92630a337eaf823218ceb658e6daf8","973d182535a01b720897f5a733a8d1b0"],"score":32.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599}],"score":28.21,"weight":20}],"eligible":true,"id":"m-f99cd5ceed3a5b41","in_window":true,"name":"Solar Mini 4","official_evidence":{"artifact_sha256":"967417099d792a2cbc0abac952f572bc529067f8ba0587e3c041a5f253bc2a0e","basis":"official_release_page","date":"2026-10-01","sha256":"967417099d792a2cbc0abac952f572bc529067f8ba0587e3c041a5f253bc2a0e","url":"https://www.upstage.ai/blog/en/solar-mini-4"},"primary_profile_id":"native","profile_id":"native","rank":76,"records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dde8616d382b2b0a447eb1104181da5e","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v2.1","peer_count":82,"percentile":40.74,"percentile_exact":{"denominator":27,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1091.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"solar-mini4:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1091.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7620c0fb57936004d0ff80edb6675f9f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"solar-mini4:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b1f8ac835c2c56f569ae7093d49e508f","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":42.11,"percentile_exact":{"denominator":19,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.22252425682190993,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"solar-mini4:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.22252425682190993},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b76daa2b8d1a11ba3f7dd4198d33e0a0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":78,"percentile":28.57,"percentile_exact":{"denominator":7,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":867.74,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"solar-mini4:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":867.74},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7e2c6c9bb5b766f6065fc5409408479","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"solar-mini4:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e92630a337eaf823218ceb658e6daf8","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18233333333333332,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18233333333333332},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Solar Mini 4","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"973d182535a01b720897f5a733a8d1b0","included":true,"model_id":"m-f99cd5ceed3a5b41","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6347329800244599,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"solar-mini4:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"solar-mini4:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6347329800244599}],"release_date":"2026-10-01","release_status":"published","release_verified":true,"result_updated_at":null,"score":33.85,"score_exact":{"denominator":3425814,"numerator":115947415},"source_id":"upstage/solar-mini4"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":19,"numerator":850},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":850},"family":"Terminal-Bench-Science","row_ids":["01d54ae905ebb77f621f6cbf30eb0930"],"score":44.74}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71}],"score":44.74,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":96,"numerator":2525},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":100},"family":"GDPval-AA","row_ids":["fcb5b9889977ea6f8824ae1c412b655a"],"score":11.11},{"benchmark_version":"4.0","exact":{"denominator":8,"numerator":175},"family":"Terminal-Bench","row_ids":["bbf181bad55ef3d8ecad294914ed6b67"],"score":21.88},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":500},"family":"AA-Briefcase","row_ids":["14b2d91b066241cbb4ba0914b53ae0a4"],"score":55.56},{"benchmark_version":"2.1","exact":{"denominator":3,"numerator":50},"family":"OSWorld","row_ids":["cdb864bc0c75bd1fb1f8c43c173d544c"],"score":16.67}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"v1.1","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0}],"score":26.3,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":125},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":125},"family":"DeepSWE","row_ids":["494d4b4ce09501b6feb0dbb80233b707"],"score":62.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0}],"score":62.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":150},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":150},"family":"SimpleQA Verified","row_ids":["0c6f13f860966ddef45cc04368fd5127"],"score":8.82}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329}],"score":8.82,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":250},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":500},"family":"Humanity Sixth Sense","row_ids":["cf4a733238c901c9de673ff40841fce0"],"score":18.52},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"Chartography","row_ids":["a4e7fdc6e1f2295b627a11265f28666d"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","peer_count":28,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6}],"score":9.26,"weight":10}],"eligible":true,"id":"m-2cd37409a483435a","in_window":true,"name":"Claude Sonnet 5","official_evidence":{"artifact_sha256":"78c9b451825b0f421c2e785a5934defd9662507c3913b2f7c2c0df07917ba424","basis":"official_release_page","date":"2026-06-30","sha256":"78c9b451825b0f421c2e785a5934defd9662507c3913b2f7c2c0df07917ba424","url":"https://www.anthropic.com/news/claude-sonnet-5"},"primary_profile_id":"native","profile_id":"native","rank":77,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"01d54ae905ebb77f621f6cbf30eb0930","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"0.1.0","peer_count":39,"percentile":44.74,"percentile_exact":{"denominator":19,"numerator":850},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":5.71,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"5.71%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row24"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row24","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":5.71},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"fcb5b9889977ea6f8824ae1c412b655a","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Knowledge work GDPval-AA v2.1³; preserve page footnotes and task-specific settings","protocol_id":"34c76587afd7902bbcf38773","raw_value":1449.0,"reference_revision":"550a2e0280c50d9ee4af0135bd6f1b8991ff975945bef1f6ba38c998c5116862","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"1449","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row9:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:GDPval-AA:Knowledge work GDPval-AA v2.1","source_record_id":"official-4e3ca8a95d36:table0:row9:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"分","value":1449.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Sonnet 5 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"bbf181bad55ef3d8ecad294914ed6b67","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"4.0","peer_count":49,"percentile":21.88,"percentile_exact":{"denominator":8,"numerator":175},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":10.61,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"10.61%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row42"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row42","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":10.61},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Sonnet 5 (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"14b2d91b066241cbb4ba0914b53ae0a4","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"v1.1","peer_count":19,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.4,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e2f5208a69231a79afde61cd"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e2f5208a69231a79afde61cd","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.4},{"benchmark_family":"OSWorld","benchmark_id":"b-66986380e1f0a0227a56","benchmark_label":"OSWorld","benchmark_version":"2.1","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"agents","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"cdb864bc0c75bd1fb1f8c43c173d544c","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"2.1","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Computer use OSWorld 2.1; preserve page footnotes and task-specific settings","protocol_id":"e6a07dd8a64cf51e658fe61b","raw_value":57.0,"reference_revision":"98eb0847e62fc75ce25b45ec3fc148b27a8716c501b3be4487e32a369713c093","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"57.0% partial","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row15:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:OSWorld:Computer use OSWorld 2.1","source_record_id":"official-4e3ca8a95d36:table0:row15:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":57.0},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:27:32.323108+00:00","component_key":"score","configuration":"Claude Sonnet 5","dimension":"coding","direction":"higher_better","feed_id":"public-release-4b569e3e16302155","id":"494d4b4ce09501b6feb0dbb80233b707","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"1.1","peer_count":5,"percentile":62.5,"percentile_exact":{"denominator":2,"numerator":125},"profile_key":"native","protocol_description":"Published page official-4b569e3e1630; row DeepSWE v1.1 Long-horizon software engineering; preserve page footnotes and task-specific settings","protocol_id":"7ed0e78158569ef8175dbb6a","raw_value":54.0,"reference_revision":"7efdf5b236a7e1f03486ed081586a98f267296b8ff38d28bdc1e2da39679cc99","research_artifact_sha256":"2b37edc98c7477174de84b274b01b679182e8404b5fa17db18af940daa64e030","response_sha256":"80d27df1b1b45a423f65ba121cf80b088707a467125b5217686ee06d804ff5e4","result_date":"2026-07-21","source_cell":"54%","source_id":"s-94bac3ae47353f5bf78d","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4b569e3e1630:table0:row4:column7"},"source_name":"deepmind.google","source_protocol_key":"official:deepmind.google:DeepSWE:DeepSWE v1.1 Long-horizon software engineering","source_record_id":"official-4b569e3e1630:table0:row4:column7","source_url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/","unit":"%","value":54.0},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-sonnet-5_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"0c6f13f860966ddef45cc04368fd5127","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"run 1.2.0","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.329,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:28:11.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"183eba1496e9a03042603b97"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"183eba1496e9a03042603b97","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.329},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Sonnet-5 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"cf4a733238c901c9de673ff40841fce0","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"current public protocol","peer_count":28,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":24.8,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"802019f50120ad156d83517d"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"802019f50120ad156d83517d","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":24.8},{"benchmark_family":"Chartography","benchmark_id":"b-94e4df78ac964870c168","benchmark_label":"Chartography","benchmark_version":"current published task suite","captured_at":"2026-11-06T04:28:18.739704+00:00","component_key":"score","configuration":"Sonnet 5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-4e3ca8a95d366330","id":"a4e7fdc6e1f2295b627a11265f28666d","included":true,"model_id":"m-2cd37409a483435a","original_version_label":"published unversioned benchmark; tool and no-tool protocols separate","peer_count":6,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page official-4e3ca8a95d36; row Visual chart recognition Chartography; preserve page footnotes and task-specific settings","protocol_id":"363770683a232ce7f97c2d56","raw_value":15.6,"reference_revision":"44f713ebd5a68e1aeae2aac20143575b9bbdf792b3f96ac99b0a2fd02d09932f","research_artifact_sha256":"95e3ecb593991a52b9c75626d0e0c3d14a2fbf0b155b225fc73e9a67ff161089","response_sha256":"80b4362d50c8847398685828367227ab762401695b0180bdac149a11e6501170","result_date":"2026-09-28","source_cell":"15.6% no tools","source_id":"s-c59f3bf1f06fd2e80e5e","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"official-4e3ca8a95d36:table0:row17:column2"},"source_name":"www.anthropic.com","source_protocol_key":"official:www.anthropic.com:Chartography:Visual chart recognition Chartography","source_record_id":"official-4e3ca8a95d36:table0:row17:column2","source_url":"https://www.anthropic.com/claude-sonnet-5-5","unit":"%","value":15.6}],"release_date":"2026-06-30","release_status":"published","release_verified":true,"result_updated_at":"2026-09-28","score":32.95,"score_exact":{"denominator":1116288,"numerator":36782095},"source_id":"anthropic/claude-sonnet-5"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":41769,"numerator":1004075},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["42505f6b2ea9972da7c497ad966aebac"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1600},"family":"HLE","row_ids":["30beb0844cc09771243cbe9dfff243cd"],"score":25.4},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":500},"family":"CritPt","row_ids":["e6e2851045fbef435d3da74124d06523"],"score":29.41}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"Diamond","peer_count":64,"percentile":25.4,"percentile_exact":{"denominator":63,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","peer_count":52,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135}],"score":24.04,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":10773,"numerator":369385},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":4000},"family":"GDPval-AA","row_ids":["3c9b5c57e54a04df82602451589cb092"],"score":49.38},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["01e4bcd3555fdf1970b9b729ff28db44"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":600},"family":"AutomationBench","row_ids":["ae26dd2850d89ddaa6b644b249715c9f"],"score":31.58},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":300},"family":"AA-Briefcase","row_ids":["b0a10dcacf57ed9fc64b8506d23f5ae1"],"score":42.86}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2.1","peer_count":82,"percentile":49.38,"percentile_exact":{"denominator":81,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":78,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76}],"score":34.29,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1600},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1600},"family":"FrontierCode","row_ids":["35228bba34f7d198ea6ae7b85139d6dc"],"score":32.65}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"2.0","peer_count":50,"percentile":32.65,"percentile_exact":{"denominator":49,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141}],"score":32.65,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":79475},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":675},"family":"GDP.pdf","row_ids":["26277d899631407b16323e0f427ca69b"],"score":35.53},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["41481c594a0d6acdcc3cd6076cd49938","1c32657b7924b6111d38086436775d4a"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161}],"score":39.46,"weight":20}],"eligible":true,"id":"m-36451e3f44503d37","in_window":true,"name":"Nemotron 3.5 Super VL","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":78,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"42505f6b2ea9972da7c497ad966aebac","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nemotron-3-5-super-vl:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"30beb0844cc09771243cbe9dfff243cd","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"Diamond","peer_count":64,"percentile":25.4,"percentile_exact":{"denominator":63,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.088,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-super-vl:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e6e2851045fbef435d3da74124d06523","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2 Hard","peer_count":52,"percentile":29.41,"percentile_exact":{"denominator":17,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.135,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nemotron-3-5-super-vl:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.135},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3c9b5c57e54a04df82602451589cb092","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v2.1","peer_count":82,"percentile":49.38,"percentile_exact":{"denominator":81,"numerator":4000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1162.41,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-super-vl:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1162.41},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"01e4bcd3555fdf1970b9b729ff28db44","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-super-vl:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ae26dd2850d89ddaa6b644b249715c9f","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14818345798643587,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-super-vl:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14818345798643587},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b0a10dcacf57ed9fc64b8506d23f5ae1","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":78,"percentile":42.86,"percentile_exact":{"denominator":7,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":998.76,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-super-vl:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":998.76},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"35228bba34f7d198ea6ae7b85139d6dc","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"2.0","peer_count":50,"percentile":32.65,"percentile_exact":{"denominator":49,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0610914141414141,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nemotron-3-5-super-vl:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0610914141414141},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"26277d899631407b16323e0f427ca69b","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"current AA protocol","peer_count":77,"percentile":35.53,"percentile_exact":{"denominator":19,"numerator":675},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.08,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-super-vl:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"41481c594a0d6acdcc3cd6076cd49938","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":22.64,"percentile_exact":{"denominator":53,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21166666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21166666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Super VL","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"1c32657b7924b6111d38086436775d4a","included":true,"model_id":"m-36451e3f44503d37","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6951374207188161,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-super-vl:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nemotron-3-5-super-vl:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6951374207188161}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"result_updated_at":null,"score":32.23,"score_exact":{"denominator":7949601387,"numerator":256194148075},"source_id":"published/aa:nemotron-3-5-super-vl"},{"benchmark_count":20,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":26078,"numerator":848075},"families":[{"benchmark_version":"2026","exact":{"denominator":3,"numerator":200},"family":"AIME","row_ids":["9bcefe5681f1de0d56a93cb5ca7a6411"],"score":66.67},{"benchmark_version":"500-item subset","exact":{"denominator":1,"numerator":100},"family":"MATH-500","row_ids":["48e7cd14f60fe5daf4244ca5ca1c19b2"],"score":100.0},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":600},"family":"GPQA Diamond","row_ids":["7e93a07b82dfd690e1c2b1309992332e"],"score":10.17},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d1c19809bb6f7e59cd9274e82fb1b4b5"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":0},"family":"HLE","row_ids":["20cab07f9ffc43544259f41d417e323c"],"score":0.0},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":50},"family":"CritPt","row_ids":["e23c22def1e7c8613f7559601fcfa581"],"score":0.98}],"family_count":6,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":10.17,"percentile_exact":{"denominator":59,"numerator":600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","peer_count":64,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":32.52,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":6162156,"numerator":60705805},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":125},"family":"tau3-Banking","row_ids":["9e32e2abcb9a8a594fb5e6aa3bcbd501"],"score":4.81},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":100},"family":"GDPval-AA","row_ids":["a9600d8248b3ea039e7316ffc060f044"],"score":3.7},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["c56ed05cd995d4a9a967c8a8edd7c23f"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":50},"family":"AutomationBench","row_ids":["de96f31cb9e7df111014b360185583e3"],"score":2.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":100},"family":"AA-Briefcase","row_ids":["caf840132854029b414465a52c19889d"],"score":1.3},{"benchmark_version":"4","exact":{"denominator":3,"numerator":100},"family":"BFCL","row_ids":["c9019a0cb657ee535803619a061368cf"],"score":33.33}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","peer_count":82,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":2.63,"percentile_exact":{"denominator":19,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":78,"percentile":1.3,"percentile_exact":{"denominator":77,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2}],"score":9.85,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":147,"numerator":5125},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["d294135a5f841df5749d74dee1cdb9e7"],"score":3.06},{"benchmark_version":"release_v6","exact":{"denominator":3,"numerator":200},"family":"LiveCodeBench","row_ids":["2f001b0be2d50fef9306d2b3f9f1569b"],"score":66.67}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9}],"score":34.86,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":477,"numerator":25225},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":3,"numerator":200},"family":"IFBench","row_ids":["a7d850e0b3d6527713784858f859aa3f"],"score":66.67},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":200},"family":"Multi-IF","row_ids":["cf2e1f5dc39c9244f47a835c7f33569d"],"score":66.67},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"GDP.pdf","row_ids":["d9c7d28e2d9651e546a84191d6ab8ec7"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"IFEval","row_ids":["55369a687291e4012c741d522a892ee5"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":50},"family":"MMLU-Pro","row_ids":["8519642492aa86c8bdd1ad71a0f84568"],"score":50.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["432ed3bb705952b8c4dcd7c697936b37","7f11b1b4129f0979e12158aab3657dcc"],"score":33.96}],"family_count":6,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509}],"score":52.88,"weight":20}],"eligible":true,"id":"m-12d2cbffeaae72d5","in_window":true,"name":"Granite 4.2 3B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":79,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"9bcefe5681f1de0d56a93cb5ca7a6411","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2026","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":83.5,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"83.5","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":83.5},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"48e7cd14f60fe5daf4244ca5ca1c19b2","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":97.0,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"97.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":97.0},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7e93a07b82dfd690e1c2b1309992332e","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":10.17,"percentile_exact":{"denominator":59,"numerator":600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.558585858585859,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.558585858585859},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d1c19809bb6f7e59cd9274e82fb1b4b5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"20cab07f9ffc43544259f41d417e323c","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Diamond","peer_count":64,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.029,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-3b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.029},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e23c22def1e7c8613f7559601fcfa581","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9e32e2abcb9a8a594fb5e6aa3bcbd501","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-3b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a9600d8248b3ea039e7316ffc060f044","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v2.1","peer_count":82,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":209.24,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":209.24},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c56ed05cd995d4a9a967c8a8edd7c23f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-3b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"de96f31cb9e7df111014b360185583e3","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":2.63,"percentile_exact":{"denominator":19,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.003767979122295127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-3b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.003767979122295127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"caf840132854029b414465a52c19889d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":78,"percentile":1.3,"percentile_exact":{"denominator":77,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":98.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-3b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":98.83},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCL v4","benchmark_version":"4","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"agents","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"c9019a0cb657ee535803619a061368cf","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v4","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCL v4; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"add40720ac2f4db6b3eb0fb3","raw_value":52.2,"reference_revision":"d014701ac6b73f60aebf97abe8d657b1b8828f9c6ed871b7cf9bb7c36fa586c1","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"52.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row32:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCL v4","source_record_id":"twelve-minicpm2-card:table0:row32:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":52.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d294135a5f841df5749d74dee1cdb9e7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"2f001b0be2d50fef9306d2b3f9f1569b","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":58.9,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"58.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":58.9},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a7d850e0b3d6527713784858f859aa3f","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":73.0,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"73.0","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row15:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-minicpm2-card:table0:row15:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":73.0},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"cf2e1f5dc39c9244f47a835c7f33569d","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":75.9,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"75.9","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":75.9},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d9c7d28e2d9651e546a84191d6ab8ec7","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"current AA protocol","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.002,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-3b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-3b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"55369a687291e4012c741d522a892ee5","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current original Google dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.7,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.7},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"granite-4.2-3B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"8519642492aa86c8bdd1ad71a0f84568","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column granite-4.2-3B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.8,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column6","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.8},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"432ed3bb705952b8c4dcd7c697936b37","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.092,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.092},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7f11b1b4129f0979e12158aab3657dcc","included":true,"model_id":"m-12d2cbffeaae72d5","original_version_label":"v1.1","peer_count":54,"percentile":67.92,"percentile_exact":{"denominator":53,"numerator":3600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6967694566813509,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6967694566813509}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.27,"score_exact":{"denominator":41274330401304,"numerator":1290617742549175},"source_id":"published/aa:granite-4-2-3b"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":117351,"numerator":3265325},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2400},"family":"GPQA Diamond","row_ids":["da90bd5db2e44d989ddf563c8ec38c6b"],"score":40.68},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["13146daedf685d8da2f4bf015457ee2c"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1300},"family":"CritPt","row_ids":["d6e7cb95527cd29a8925ca94f594e568"],"score":25.49}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":40.68,"percentile_exact":{"denominator":59,"numerator":2400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","peer_count":52,"percentile":25.49,"percentile_exact":{"denominator":51,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08}],"score":27.83,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":24453,"numerator":1164503},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":875},"family":"tau3-Banking","row_ids":["7e3c134e84deebeefe2308a65aef1578"],"score":67.31},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":500},"family":"GDPval-AA","row_ids":["7f3c9ed4efe6520cbbf9a35bd5a39f5e"],"score":55.56},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["cd7e16f694d6c0543ca83b3b5dd9358b"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":900},"family":"AutomationBench","row_ids":["908bb98659898bf60c101cff97ce69b8"],"score":47.37},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":600},"family":"AA-Briefcase","row_ids":["54e0dbe28b524361d2c39e892adaa185"],"score":54.55}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":67.31,"percentile_exact":{"denominator":13,"numerator":875},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","peer_count":82,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":47.37,"percentile_exact":{"denominator":19,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","peer_count":78,"percentile":54.55,"percentile_exact":{"denominator":11,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79}],"score":47.62,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":1200},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1200},"family":"FrontierCode","row_ids":["33a3261e51180ab895a82ccd6f7e12a3"],"score":24.49}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","peer_count":50,"percentile":24.49,"percentile_exact":{"denominator":49,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444}],"score":24.49,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":825},"family":"GDP.pdf","row_ids":["14efb36fead599ff0b956fa101f23a86"],"score":21.71}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042}],"score":21.71,"weight":20}],"eligible":true,"id":"m-a67d9ce95a33f246","in_window":true,"name":"K2 Horizon MoVA 36B A4B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":80,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"da90bd5db2e44d989ddf563c8ec38c6b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":40.68,"percentile_exact":{"denominator":59,"numerator":2400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.822222222222222,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-mova-36b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.822222222222222},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13146daedf685d8da2f4bf015457ee2c","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d6e7cb95527cd29a8925ca94f594e568","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2 Hard","peer_count":52,"percentile":25.49,"percentile_exact":{"denominator":51,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.08,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.08},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7e3c134e84deebeefe2308a65aef1578","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":67.31,"percentile_exact":{"denominator":13,"numerator":875},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.319587628865979,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-mova-36b-a4b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.319587628865979},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7f3c9ed4efe6520cbbf9a35bd5a39f5e","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v2.1","peer_count":82,"percentile":55.56,"percentile_exact":{"denominator":9,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1196.67,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-mova-36b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1196.67},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cd7e16f694d6c0543ca83b3b5dd9358b","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-mova-36b-a4b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"908bb98659898bf60c101cff97ce69b8","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":47.37,"percentile_exact":{"denominator":19,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.2610203199558987,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-mova-36b-a4b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2610203199558987},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"54e0dbe28b524361d2c39e892adaa185","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"v1.1","peer_count":78,"percentile":54.55,"percentile_exact":{"denominator":11,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1164.79,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-mova-36b-a4b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1164.79},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"33a3261e51180ab895a82ccd6f7e12a3","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"2.0","peer_count":50,"percentile":24.49,"percentile_exact":{"denominator":49,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0450194444444444,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0450194444444444},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon MoVA 36B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"14efb36fead599ff0b956fa101f23a86","included":true,"model_id":"m-a67d9ce95a33f246","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-mova-36b-a4b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-mova-36b-a4b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.22,"score_exact":{"denominator":3605374773,"numerator":112575457730},"source_id":"published/aa:k2-horizon-mova-36b-a4b"},{"benchmark_count":6,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":1400},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1400},"family":"GPQA Diamond","row_ids":["bbee1bae6aa788cf47294ddc38f9761f"],"score":23.73}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":23.73,"percentile_exact":{"denominator":59,"numerator":1400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333}],"score":23.73,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":648648,"numerator":5385785},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":125},"family":"tau3-Banking","row_ids":["24158d9869238dabbd834ff7fd267dc4"],"score":4.81},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":800},"family":"GDPval-AA","row_ids":["fc44351a5b9d8e210a763f56506c0719"],"score":9.88},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["2b6eac192b734b825e56a7c642b341b7"],"score":13.33},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":400},"family":"AA-Briefcase","row_ids":["606c6cc32d7931469129472a4d6fffcd"],"score":5.19}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","peer_count":82,"percentile":9.88,"percentile_exact":{"denominator":81,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","peer_count":78,"percentile":5.19,"percentile_exact":{"denominator":77,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41}],"score":8.3,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4,"numerator":275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":275},"family":"IFBench","row_ids":["ea00f82eb447a348674a4f2b40a4a4f4"],"score":68.75}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136}],"score":68.75,"weight":20}],"eligible":true,"id":"m-7b2a8fe101678ebd","in_window":true,"name":"HyperNova 60B 2605","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-26","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":81,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bbee1bae6aa788cf47294ddc38f9761f","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":23.73,"percentile_exact":{"denominator":59,"numerator":1400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.733333333333333,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"hypernova-60b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.733333333333333},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"24158d9869238dabbd834ff7fd267dc4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":4.81,"percentile_exact":{"denominator":26,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0556701030927835,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"hypernova-60b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0556701030927835},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fc44351a5b9d8e210a763f56506c0719","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v2.1","peer_count":82,"percentile":9.88,"percentile_exact":{"denominator":81,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":448.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"hypernova-60b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":448.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2b6eac192b734b825e56a7c642b341b7","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"hypernova-60b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"606c6cc32d7931469129472a4d6fffcd","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"v1.1","peer_count":78,"percentile":5.19,"percentile_exact":{"denominator":77,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":232.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"hypernova-60b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":232.41},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"HyperNova 60B 2605 (High, Based on gpt-oss-120b)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ea00f82eb447a348674a4f2b40a4a4f4","included":true,"model_id":"m-7b2a8fe101678ebd","original_version_label":"current 294-item protocol","peer_count":17,"percentile":68.75,"percentile_exact":{"denominator":4,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.664625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"hypernova-60b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"hypernova-60b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.664625850340136}],"release_date":"2026-05-26","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.08,"score_exact":{"denominator":535783248,"numerator":16653656375},"source_id":"published/aa:hypernova-60b"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1095276,"numerator":37020775},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3700},"family":"GPQA Diamond","row_ids":["198f62e91fe9f1a3e9e87c6a5a19a9fe"],"score":62.71},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["70737e236c952ff057602f3190f37b75"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":500},"family":"HLE","row_ids":["85bd1df1c636ca5159d1bed6e1a0f499"],"score":23.81},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1600},"family":"CritPt","row_ids":["f5fbcfce1c679af1818579e2368d3d27"],"score":31.37}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":62.71,"percentile_exact":{"denominator":59,"numerator":3700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","peer_count":64,"percentile":23.81,"percentile_exact":{"denominator":21,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","peer_count":52,"percentile":31.37,"percentile_exact":{"denominator":51,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15}],"score":33.8,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":10783773,"numerator":221054348},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":300},"family":"tau3-Banking","row_ids":["5c824422e02d6ae8e8e0e7805866dd38"],"score":23.08},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2800},"family":"GDPval-AA","row_ids":["d9fac2d0e8fa2fae504adca322b42dfa"],"score":34.57},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["22a0f772949251594cd3a25e02ef9752"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":350},"family":"AutomationBench","row_ids":["b55d26365439e3d2ed6eba908f339027"],"score":18.42},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2300},"family":"AA-Briefcase","row_ids":["e12061e69225a05e579dff8f6ab0f0de"],"score":29.87},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":50},"family":"EnterpriseOps-Gym","row_ids":["4c1c120149f7031df34c6f3ac8ff9629"],"score":5.56},{"benchmark_version":"current AA pass^5","exact":{"denominator":1,"numerator":0},"family":"AA-AnalystAgent","row_ids":["2275301a736d0fe2b34ce5c990945f54"],"score":0.0}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","peer_count":82,"percentile":34.57,"percentile_exact":{"denominator":81,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":78,"percentile":29.87,"percentile_exact":{"denominator":77,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","peer_count":24,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625}],"score":20.5,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":7,"numerator":100},"families":[{"benchmark_version":"2.0","exact":{"denominator":7,"numerator":100},"family":"FrontierCode","row_ids":["07260be02cf7c09201c1af77d371a386"],"score":14.29}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","peer_count":50,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576}],"score":14.29,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":232025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":375},"family":"IFBench","row_ids":["74e2baf55ca27065ec7aeb8f6eee9414"],"score":93.75},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":925},"family":"GDP.pdf","row_ids":["994833badab5af26add99f20244353b3"],"score":24.34},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2900},"family":"AA-Omniscience","row_ids":["7a931d4238635511e69e1feb35bfd055","f659a47d7f8491b1073ef450bc951aa0"],"score":54.72}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","peer_count":17,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645}],"score":57.6,"weight":20}],"eligible":true,"id":"m-ad7decc9ae94df12","in_window":true,"name":"Nemotron 3 Ultra 550B A55B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-04","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":82,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"198f62e91fe9f1a3e9e87c6a5a19a9fe","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":62.71,"percentile_exact":{"denominator":59,"numerator":3700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.866666666666667,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.866666666666667},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70737e236c952ff057602f3190f37b75","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"85bd1df1c636ca5159d1bed6e1a0f499","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"Diamond","peer_count":64,"percentile":23.81,"percentile_exact":{"denominator":21,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.085,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.085},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f5fbcfce1c679af1818579e2368d3d27","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2 Hard","peer_count":52,"percentile":31.37,"percentile_exact":{"denominator":51,"numerator":1600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.15,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.15},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5c824422e02d6ae8e8e0e7805866dd38","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":23.08,"percentile_exact":{"denominator":13,"numerator":300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.142268041237113,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.142268041237113},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9fac2d0e8fa2fae504adca322b42dfa","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v2.1","peer_count":82,"percentile":34.57,"percentile_exact":{"denominator":81,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1020.69,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1020.69},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22a0f772949251594cd3a25e02ef9752","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b55d26365439e3d2ed6eba908f339027","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.02986349541593992,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.02986349541593992},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e12061e69225a05e579dff8f6ab0f0de","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":78,"percentile":29.87,"percentile_exact":{"denominator":77,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":877.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":877.09},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4c1c120149f7031df34c6f3ac8ff9629","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA run","peer_count":19,"percentile":5.56,"percentile_exact":{"denominator":9,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.28886899433005075,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.28886899433005075},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2275301a736d0fe2b34ce5c990945f54","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA pass^5","peer_count":24,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.0625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0625},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"07260be02cf7c09201c1af77d371a386","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"2.0","peer_count":50,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0109007575757576,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0109007575757576},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"74e2baf55ca27065ec7aeb8f6eee9414","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current 294-item protocol","peer_count":17,"percentile":93.75,"percentile_exact":{"denominator":4,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.813605442176871,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.813605442176871},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"994833badab5af26add99f20244353b3","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"current AA protocol","peer_count":77,"percentile":24.34,"percentile_exact":{"denominator":38,"numerator":925},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.05,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.05},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7a931d4238635511e69e1feb35bfd055","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.21233333333333335,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.21233333333333335},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3 Ultra 550B A55B (Reasoning)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"f659a47d7f8491b1073ef450bc951aa0","included":true,"model_id":"m-ad7decc9ae94df12","original_version_label":"v1.1","peer_count":54,"percentile":84.91,"percentile_exact":{"denominator":53,"numerator":4500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.774016081252645,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"nvidia-nemotron-3-ultra-550b-a55b:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.774016081252645}],"release_date":"2026-06-04","release_status":"published","release_verified":true,"result_updated_at":null,"score":31.06,"score_exact":{"denominator":41274330401304,"numerator":1281914164344215},"source_id":"nvidia/nemotron-3-ultra-550b-a55b"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":177,"numerator":500},"families":[{"benchmark_version":"2026","exact":{"denominator":1,"numerator":0},"family":"AIME","row_ids":["1aa7620d5715919ff1d95bb835bf5bb1"],"score":0.0},{"benchmark_version":"500-item subset","exact":{"denominator":1,"numerator":0},"family":"MATH-500","row_ids":["d2a6fbf18a7d28ad26aab8f6363bd884"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":500},"family":"GPQA Diamond","row_ids":["c0382c8dd0c506272b3fef2eaf1a7bcf"],"score":8.47}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":8.47,"percentile_exact":{"denominator":59,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758}],"score":2.82,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3159,"numerator":120050},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":150},"family":"tau3-Banking","row_ids":["ad5761dce64ad5d75bd8fe071f2c9120"],"score":11.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":200},"family":"GDPval-AA","row_ids":["7d56dec4d055d3d9b1237e11f0bf7ef4"],"score":2.47},{"benchmark_version":"4","exact":{"denominator":1,"numerator":100},"family":"BFCL","row_ids":["78759a40401be7663e3957b80c649826"],"score":100.0}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","peer_count":82,"percentile":2.47,"percentile_exact":{"denominator":81,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88}],"score":38.0,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3,"numerator":100},"families":[{"benchmark_version":"release_v6","exact":{"denominator":3,"numerator":100},"family":"LiveCodeBench","row_ids":["ff54ed0b44635ad5d8a708f3d17d4063"],"score":33.33}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1}],"score":33.33,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":36,"numerator":1825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"IFBench","row_ids":["49b7d39b3c4a3eaca06d3883d0335282"],"score":11.11},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":100},"family":"Multi-IF","row_ids":["ddcaace3bb0b4a1452f5dea314dd058c"],"score":100.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":200},"family":"IFEval","row_ids":["a24e132d4558f677a1ce0a399471dbb9"],"score":66.67},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":25},"family":"MMLU-Pro","row_ids":["239f2b6a72085e3eb63d599375b5b516"],"score":25.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2}],"score":50.69,"weight":20}],"eligible":true,"id":"m-d0874ed91fd69e8e","in_window":true,"name":"LFM2.5-2.6B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-04","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":83,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"1aa7620d5715919ff1d95bb835bf5bb1","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"2026","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":45.2,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"45.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":45.2},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"d2a6fbf18a7d28ad26aab8f6363bd884","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":89.6,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"89.6","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":89.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0382c8dd0c506272b3fef2eaf1a7bcf","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":8.47,"percentile_exact":{"denominator":59,"numerator":500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.557575757575758,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-2-6b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.557575757575758},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ad5761dce64ad5d75bd8fe071f2c9120","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":11.54,"percentile_exact":{"denominator":13,"numerator":150},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0721649484536082,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"lfm2-5-2-6b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0721649484536082},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d56dec4d055d3d9b1237e11f0bf7ef4","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v2.1","peer_count":82,"percentile":2.47,"percentile_exact":{"denominator":81,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-0.53,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-2-6b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"lfm2-5-2-6b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-0.53},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"agents","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"78759a40401be7663e3957b80c649826","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"v4","peer_count":2,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":56.88,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"56.88","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row9:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid26-card:table4:row9:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":56.88},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"ff54ed0b44635ad5d8a708f3d17d4063","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":42.1,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"42.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":42.1},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"49b7d39b3c4a3eaca06d3883d0335282","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current release dataset; evaluator-specific strict/loose protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFBench; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":59.17,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"59.17","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row6:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"twelve-liquid26-card:table4:row6:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":59.17},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:24.150819+00:00","component_key":"score","configuration":"LFM2.5-2.6B (2.6B)","dimension":"language_information","direction":"higher_better","feed_id":"public-release-be95e4c0c1fd5cd2","id":"ddcaace3bb0b4a1452f5dea314dd058c","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":100.0,"percentile_exact":{"denominator":1,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-2.6B (2.6B). Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":80.07,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"12ea5a0ffff643cbd71b6f2e7a2ab253398f2c050812c4b695b11a1c72f7e61a","response_sha256":"e9fc0186733b545c15794f1c09bc5d637bef9e8e07c2fd2bf8e5089a1c68aa5c","result_date":null,"source_cell":"80.07","source_id":"s-095595bcbc77f3c90537","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid26-card:table4:row7:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-liquid26-card:table4:row7:column1","source_url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B","unit":"%","value":80.07},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"a24e132d4558f677a1ce0a399471dbb9","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current original Google dataset","peer_count":4,"percentile":66.67,"percentile_exact":{"denominator":3,"numerator":200},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":93.4,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.4},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-2.6B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"239f2b6a72085e3eb63d599375b5b516","included":true,"model_id":"m-d0874ed91fd69e8e","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-2.6B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":65.2,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"65.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column2","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":65.2}],"release_date":"2026-08-04","release_status":"published","release_verified":true,"result_updated_at":null,"score":30.01,"score_exact":{"denominator":3354858,"numerator":100691975},"source_id":"published/aa:lfm2-5-2-6b"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":123000875},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":3800},"family":"GPQA Diamond","row_ids":["a95cdd62235bf41ff0b78b6a5a866d88"],"score":64.41},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["8120a268fd2b22fe2a65df299e1ab0c5"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":2000},"family":"HLE","row_ids":["2d291e3811085b2648de857339509203"],"score":31.75},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":1850},"family":"CritPt","row_ids":["a9467dab432c3fd1ab9b333380c66f83"],"score":36.27}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":64.41,"percentile_exact":{"denominator":59,"numerator":3800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","included":true,"model_id":"m-b439c804161fac26","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond","peer_count":64,"percentile":31.75,"percentile_exact":{"denominator":63,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18}],"score":37.43,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":70864794,"numerator":2278099781},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":850},"family":"tau3-Banking","row_ids":["22b4395fa3e0e77e06e9fcfe630f8789"],"score":65.38},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3200},"family":"GDPval-AA","row_ids":["ff011af995e95f243ad573421cc4ba05"],"score":39.51},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["1344e9967afa11034830d610f592ed83"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":450},"family":"AutomationBench","row_ids":["5bd1e6ef9a7eb6f30e6c6bf694074f2d"],"score":23.68},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2000},"family":"AA-Briefcase","row_ids":["7d5fcb038f359de5556a57151b51cffc"],"score":25.97},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":350},"family":"EnterpriseOps-Gym","row_ids":["40747d0e01fbdeb26f3ad425860f13bb"],"score":38.89},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":0},"family":"MCP Atlas","row_ids":["73ec2c70e1808cc25e2b747786aa5d99"],"score":0.0},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":500},"family":"AA-AnalystAgent","row_ids":["753dbaee803264fefa51f263a84b1509"],"score":21.74}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","included":true,"model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2.1","peer_count":82,"percentile":39.51,"percentile_exact":{"denominator":81,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":23.68,"percentile_exact":{"denominator":19,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":78,"percentile":25.97,"percentile_exact":{"denominator":77,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA run","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","peer_count":24,"percentile":21.74,"percentile_exact":{"denominator":23,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375}],"score":32.15,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":49,"numerator":550},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1100},"family":"FrontierCode","row_ids":["9b48c9ab902cc3f53ca099a627db7e76"],"score":22.45},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"Vals Code Migration","row_ids":["2bad0033060ca0f870378d7a7d3d6c5d"],"score":0.0}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","included":true,"model_id":"m-b439c804161fac26","original_version_label":"2.0","peer_count":50,"percentile":22.45,"percentile_exact":{"denominator":49,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1.1","peer_count":27,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82}],"score":11.22,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":51357,"numerator":2115350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":1150},"family":"GDP.pdf","row_ids":["af7da6ed00064e0d491020c30636379c"],"score":60.53},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":350},"family":"SimpleQA Verified","row_ids":["28411df2fb9e1660a30498ca8b0c9245"],"score":20.59},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2250},"family":"AA-Omniscience","row_ids":["95037c24ac83fd0eadfbf39605ac569f","3698e24fd77dd4f8bc27edd44ed2efc7"],"score":42.45}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","included":true,"model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329}],"score":41.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":350},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":350},"family":"MMMU-Pro","row_ids":["874b32395a8c924c1d3f0faf3d2e2b2e"],"score":20.59}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856}],"score":20.59,"weight":10}],"eligible":true,"id":"m-b439c804161fac26","in_window":true,"name":"Inkling","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":84,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a95cdd62235bf41ff0b78b6a5a866d88","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":64.41,"percentile_exact":{"denominator":59,"numerator":3800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.871717171717172,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.871717171717172},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8120a268fd2b22fe2a65df299e1ab0c5","included":true,"model_id":"m-b439c804161fac26","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2d291e3811085b2648de857339509203","included":true,"model_id":"m-b439c804161fac26","original_version_label":"Diamond","peer_count":64,"percentile":31.75,"percentile_exact":{"denominator":63,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.106,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.106},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"a9467dab432c3fd1ab9b333380c66f83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2 Hard","peer_count":52,"percentile":36.27,"percentile_exact":{"denominator":51,"numerator":1850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.18,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"22b4395fa3e0e77e06e9fcfe630f8789","included":true,"model_id":"m-b439c804161fac26","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":65.38,"percentile_exact":{"denominator":13,"numerator":850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.290721649484536,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.290721649484536},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ff011af995e95f243ad573421cc4ba05","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v2.1","peer_count":82,"percentile":39.51,"percentile_exact":{"denominator":81,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1080.22,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1080.22},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1344e9967afa11034830d610f592ed83","included":true,"model_id":"m-b439c804161fac26","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"inkling:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5bd1e6ef9a7eb6f30e6c6bf694074f2d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":23.68,"percentile_exact":{"denominator":19,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.04984889684475201,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04984889684475201},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7d5fcb038f359de5556a57151b51cffc","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":78,"percentile":25.97,"percentile_exact":{"denominator":77,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":833.05,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":833.05},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"40747d0e01fbdeb26f3ad425860f13bb","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA run","peer_count":19,"percentile":38.89,"percentile_exact":{"denominator":9,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.37958818263205013,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37958818263205013},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling (xHigh)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"73ec2c70e1808cc25e2b747786aa5d99","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":76,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d88b105ebe1f55640c5773dd"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d88b105ebe1f55640c5773dd","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":76},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"753dbaee803264fefa51f263a84b1509","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA pass^5","peer_count":24,"percentile":21.74,"percentile_exact":{"denominator":23,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2375,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2375},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"9b48c9ab902cc3f53ca099a627db7e76","included":true,"model_id":"m-b439c804161fac26","original_version_label":"2.0","peer_count":50,"percentile":22.45,"percentile_exact":{"denominator":49,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0392487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0392487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"2bad0033060ca0f870378d7a7d3d6c5d","included":true,"model_id":"m-b439c804161fac26","original_version_label":"1.1","peer_count":27,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":8.82,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"8.82%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row28"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row28","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":8.82},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"af7da6ed00064e0d491020c30636379c","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current AA protocol","peer_count":77,"percentile":60.53,"percentile_exact":{"denominator":19,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.128,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.128},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"Inkling_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"28411df2fb9e1660a30498ca8b0c9245","included":true,"model_id":"m-b439c804161fac26","original_version_label":"run 1.2.0","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.403,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:45.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"380cc5130a221ef55c7db6fe"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"380cc5130a221ef55c7db6fe","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.403},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"95037c24ac83fd0eadfbf39605ac569f","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":64.15,"percentile_exact":{"denominator":53,"numerator":3400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.393,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.393},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling (Xhigh)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3698e24fd77dd4f8bc27edd44ed2efc7","included":true,"model_id":"m-b439c804161fac26","original_version_label":"v1.1","peer_count":54,"percentile":20.75,"percentile_exact":{"denominator":53,"numerator":1100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3278418451400329,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3278418451400329},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling (Xhigh)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"874b32395a8c924c1d3f0faf3d2e2b2e","included":true,"model_id":"m-b439c804161fac26","original_version_label":"current Standard 10","peer_count":35,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.734682080924856,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734682080924856}],"release_date":"2026-07-15","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":29.94,"score_exact":{"denominator":16227514516752,"numerator":485799335327597},"source_id":"thinkingmachines/inkling"},{"benchmark_count":18,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":365092,"numerator":15288225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":4300},"family":"GPQA Diamond","row_ids":["70d4757947a89f4f143b905fc208d5b4"],"score":72.88},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["569c655345f8f6c2c6fa591358c80baa"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":800},"family":"HLE","row_ids":["3c34699bd1ded8eb4f916044a6a266ef"],"score":38.1},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":2000},"family":"CritPt","row_ids":["0bfa1e021f0f3d031ecf9fe9a047c9e1"],"score":39.22}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":72.88,"percentile_exact":{"denominator":59,"numerator":4300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","peer_count":64,"percentile":38.1,"percentile_exact":{"denominator":21,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","peer_count":52,"percentile":39.22,"percentile_exact":{"denominator":51,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185}],"score":41.87,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":755891136,"numerator":23344506125},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":525},"family":"tau3-Banking","row_ids":["b9bbbaa648034918f0dabaffc06da3bd"],"score":40.38},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":1300},"family":"GDPval-AA","row_ids":["fd7e6b8b2243df420c1327f60f09d326"],"score":48.15},{"benchmark_version":"4.0","exact":{"denominator":24,"numerator":25},"family":"Terminal-Bench","row_ids":["2f300304d9d00cf74194cf028a1e2a05"],"score":1.04},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":425},"family":"AutomationBench","row_ids":["441205ca0d2d0b6e786d210564b45527"],"score":22.37},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2500},"family":"AA-Briefcase","row_ids":["b87f8c47a158b30537b60e5453c5b4d9"],"score":32.47},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":550},"family":"EnterpriseOps-Gym","row_ids":["fe3cd8a1335112d2ed6891e37698119e"],"score":61.11},{"benchmark_version":"1000 tasks","exact":{"denominator":9,"numerator":100},"family":"MCP Atlas","row_ids":["72495a708e9f00af5b2d17370ac07628"],"score":11.11},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":700},"family":"AA-AnalystAgent","row_ids":["6890e628f5988bd5a798a680de5a69d4"],"score":30.43}],"family_count":8,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":40.38,"percentile_exact":{"denominator":13,"numerator":525},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","peer_count":82,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"4.0","peer_count":49,"percentile":1.04,"percentile_exact":{"denominator":24,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":22.37,"percentile_exact":{"denominator":19,"numerator":425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":78,"percentile":32.47,"percentile_exact":{"denominator":77,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","peer_count":24,"percentile":30.43,"percentile_exact":{"denominator":23,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275}],"score":30.88,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":637,"numerator":9675},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1300},"family":"FrontierCode","row_ids":["066fdac9a6f7fc061fce980c5a691276"],"score":26.53},{"benchmark_version":"1.1","exact":{"denominator":13,"numerator":50},"family":"Vals Code Migration","row_ids":["17347953383a652f8a8fabeabf1d07e6"],"score":3.85}],"family_count":2,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"2.0","peer_count":50,"percentile":26.53,"percentile_exact":{"denominator":49,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1.1","peer_count":27,"percentile":3.85,"percentile_exact":{"denominator":13,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43}],"score":15.19,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":2,"exact":{"denominator":6042,"numerator":170825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1825},"family":"GDP.pdf","row_ids":["17242a9341b3cc9173ea459907ce9455"],"score":48.03},{"benchmark_version":"run 1.2.0","exact":{"denominator":1,"numerator":0},"family":"SimpleQA Verified","row_ids":["62e95547f0f7073efcdd1420b882e4a5"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1950},"family":"AA-Omniscience","row_ids":["c31ed6f5932783f10dbc8e7e910353cc","01245c922f864bcfcfa4ad1336dea016"],"score":36.79}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433}],"score":28.27,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":400},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":400},"family":"MMMU-Pro","row_ids":["93ce1046100f7e346ed0a342303a4604"],"score":23.53}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665}],"score":23.53,"weight":10}],"eligible":true,"id":"m-35c4f287acc76c95","in_window":true,"name":"Inkling Small","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-30","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":85,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"70d4757947a89f4f143b905fc208d5b4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":72.88,"percentile_exact":{"denominator":59,"numerator":4300},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.894949494949495,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"inkling-small:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.894949494949495},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"569c655345f8f6c2c6fa591358c80baa","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"inkling-small:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3c34699bd1ded8eb4f916044a6a266ef","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"Diamond","peer_count":64,"percentile":38.1,"percentile_exact":{"denominator":21,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.121,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"inkling-small:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.121},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"0bfa1e021f0f3d031ecf9fe9a047c9e1","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2 Hard","peer_count":52,"percentile":39.22,"percentile_exact":{"denominator":51,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.185,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"inkling-small:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.185},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b9bbbaa648034918f0dabaffc06da3bd","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":40.38,"percentile_exact":{"denominator":13,"numerator":525},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.187628865979381,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"inkling-small:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.187628865979381},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fd7e6b8b2243df420c1327f60f09d326","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v2.1","peer_count":82,"percentile":48.15,"percentile_exact":{"denominator":27,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1132.56,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"inkling-small:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1132.56},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Inkling Small / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2f300304d9d00cf74194cf028a1e2a05","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"4.0","peer_count":49,"percentile":1.04,"percentile_exact":{"denominator":24,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"0.00%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row56"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row56","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":0.0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"441205ca0d2d0b6e786d210564b45527","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":22.37,"percentile_exact":{"denominator":19,"numerator":425},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.048142275152503786,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"inkling-small:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.048142275152503786},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87f8c47a158b30537b60e5453c5b4d9","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":78,"percentile":32.47,"percentile_exact":{"denominator":77,"numerator":2500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":912.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"inkling-small:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":912.63},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"fe3cd8a1335112d2ed6891e37698119e","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA run","peer_count":19,"percentile":61.11,"percentile_exact":{"denominator":9,"numerator":550},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.4270367054610564,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"inkling-small:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.4270367054610564},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"Inkling-small","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"72495a708e9f00af5b2d17370ac07628","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":79.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"7d276d439d0c42dd69648ef1"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"7d276d439d0c42dd69648ef1","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":79.2},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6890e628f5988bd5a798a680de5a69d4","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA pass^5","peer_count":24,"percentile":30.43,"percentile_exact":{"denominator":23,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.275,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"inkling-small:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.275},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"066fdac9a6f7fc061fce980c5a691276","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"2.0","peer_count":50,"percentile":26.53,"percentile_exact":{"denominator":49,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0477487373737374,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"inkling-small:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0477487373737374},{"benchmark_family":"Vals Code Migration","benchmark_id":"b-33e41e23ffa8c9de8b28","benchmark_label":"Vals Code Migration","benchmark_version":"1.1","captured_at":"2026-11-08T03:28:53.531874+00:00","component_key":"score","configuration":"Inkling Small / Published Code Migration 1.1 sandbox","dimension":"coding","direction":"higher_better","feed_id":"public-vals-valscodemigration","id":"17347953383a652f8a8fabeabf1d07e6","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"1.1","peer_count":27,"percentile":3.85,"percentile_exact":{"denominator":13,"numerator":50},"profile_key":"native","protocol_description":"Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"08d602c28ff5aa288b184edb","raw_value":10.43,"reference_revision":"74faf24ee58fca9076cd7925500491c2b21a4ba583dc0de80bf8a6c33e60a927","research_artifact_sha256":"99b6546cd7649c2feeb83cb8c9a8342dca9f923042525d5e57db9be267368ab5","response_sha256":"68f65d1638726546edffb0587d7263ed7eb0ff2c8c7953559f3466f2fcbbca4a","result_date":null,"source_cell":"10.43%","source_id":"s-2dc8e77e0b1e357b6075","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-migration11:row27"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Published Code Migration 1.1 sandbox. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-migration11:row27","source_url":"https://www.vals.ai/benchmarks/code-migration-v1-1","unit":"%","value":10.43},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"17242a9341b3cc9173ea459907ce9455","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current AA protocol","peer_count":77,"percentile":48.03,"percentile_exact":{"denominator":38,"numerator":1825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.112,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"inkling-small:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.112},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"inkling-small_xhigh","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"62e95547f0f7073efcdd1420b882e4a5","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"run 1.2.0","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.191,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:17.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"def538b5109e3af8f9450274"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"def538b5109e3af8f9450274","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.191},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c31ed6f5932783f10dbc8e7e910353cc","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":49.06,"percentile_exact":{"denominator":53,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.318,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.318},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Inkling Small","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"01245c922f864bcfcfa4ad1336dea016","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"v1.1","peer_count":54,"percentile":24.53,"percentile_exact":{"denominator":53,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.3949169110459433,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"inkling-small:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3949169110459433},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Inkling Small","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"93ce1046100f7e346ed0a342303a4604","included":true,"model_id":"m-35c4f287acc76c95","original_version_label":"current Standard 10","peer_count":35,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.740462427745665,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"inkling-small:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"inkling-small:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.740462427745665}],"release_date":"2026-07-30","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":29.23,"score_exact":{"denominator":1125107673161472,"numerator":32892305162137165},"source_id":"thinkingmachines/inkling-small"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":177,"numerator":6250},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2200},"family":"GPQA Diamond","row_ids":["7c995d56b5fa39eef927afa2cc879ab9"],"score":37.29},{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":100},"family":"HLE","row_ids":["12e6c7bd6a0e1d4d1047be66dc2b360a"],"score":33.33}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":37.29,"percentile_exact":{"denominator":59,"numerator":2200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond","peer_count":64,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115}],"score":35.31,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":32683129},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":275},"family":"tau3-Banking","row_ids":["69ddb67fade6ab2b7cbae1dad6474f61"],"score":21.15},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2200},"family":"GDPval-AA","row_ids":["d1bad78363493513c436f20f24fa1380"],"score":27.16},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["27a6f0107d1b80cc1605307626b6cc35"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":400},"family":"AutomationBench","row_ids":["5cc05f32d32a7a675fc47d1b76ed156a"],"score":21.05},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1800},"family":"AA-Briefcase","row_ids":["1e5075fa69ac15167616fbc24b2c2ddf"],"score":23.38}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":21.15,"percentile_exact":{"denominator":13,"numerator":275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v2.1","peer_count":82,"percentile":27.16,"percentile_exact":{"denominator":81,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":21.05,"percentile_exact":{"denominator":19,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v1.1","peer_count":78,"percentile":23.38,"percentile_exact":{"denominator":77,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73}],"score":21.22,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1175},"family":"GDP.pdf","row_ids":["c1a1527152276d41667adbe5c22b811f"],"score":30.92}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"score":30.92,"weight":20}],"eligible":true,"id":"m-6ac02cf910375655","in_window":true,"name":"LongCat 2.0","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-29","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":86,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"7c995d56b5fa39eef927afa2cc879ab9","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":37.29,"percentile_exact":{"denominator":59,"numerator":2200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.77979797979798,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"longcat-2-0:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.77979797979798},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"12e6c7bd6a0e1d4d1047be66dc2b360a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"Diamond","peer_count":64,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.115,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"longcat-2-0:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"69ddb67fade6ab2b7cbae1dad6474f61","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":21.15,"percentile_exact":{"denominator":13,"numerator":275},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.131958762886598,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"longcat-2-0:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.131958762886598},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d1bad78363493513c436f20f24fa1380","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v2.1","peer_count":82,"percentile":27.16,"percentile_exact":{"denominator":81,"numerator":2200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":876.64,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"longcat-2-0:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":876.64},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"27a6f0107d1b80cc1605307626b6cc35","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"longcat-2-0:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5cc05f32d32a7a675fc47d1b76ed156a","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":21.05,"percentile_exact":{"denominator":19,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.045339131570294025,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"longcat-2-0:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045339131570294025},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1e5075fa69ac15167616fbc24b2c2ddf","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"v1.1","peer_count":78,"percentile":23.38,"percentile_exact":{"denominator":77,"numerator":1800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":739.73,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"longcat-2-0:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":739.73},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LongCat 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"c1a1527152276d41667adbe5c22b811f","included":true,"model_id":"m-6ac02cf910375655","original_version_label":"current AA protocol","peer_count":77,"percentile":30.92,"percentile_exact":{"denominator":38,"numerator":1175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.064,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"longcat-2-0:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"longcat-2-0:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.064}],"release_date":"2026-06-29","release_status":"published","release_verified":true,"result_updated_at":null,"score":29.02,"score_exact":{"denominator":1272485214,"numerator":36930684955},"source_id":"meituan/longcat-2.0"},{"benchmark_count":15,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":96088925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2900},"family":"GPQA Diamond","row_ids":["ff2b5baedd439ff80001794125500940"],"score":49.15},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["f04dc7e5afab13c2812e16f8a7f09157"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1700},"family":"HLE","row_ids":["67c349c9897445ce219d7f8acc0e4dbe"],"score":26.98},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":400},"family":"CritPt","row_ids":["142184a428fc22f5f377c4fcd96f6a10"],"score":23.53}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":49.15,"percentile_exact":{"denominator":59,"numerator":2900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","peer_count":64,"percentile":26.98,"percentile_exact":{"denominator":63,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","peer_count":52,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07}],"score":29.24,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":22547889,"numerator":644722264},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":750},"family":"tau3-Banking","row_ids":["4a499df1f84b1eaf0a172cfe97a95f4c"],"score":57.69},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2000},"family":"GDPval-AA","row_ids":["cab0e8633e6425575d15983329eb67d3"],"score":24.69},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["af8dab2781940ab31c2449f02ac9e502"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":500},"family":"AutomationBench","row_ids":["af88768655cfd944c0cc2a22a6f5661c"],"score":26.32},{"benchmark_version":"1.1","exact":{"denominator":7,"numerator":100},"family":"AA-Briefcase","row_ids":["6b81d5a0603a102ca44c8276f19acede"],"score":14.29},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":250},"family":"EnterpriseOps-Gym","row_ids":["6751eebce76aacc484338eb2b0237376"],"score":27.78},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":400},"family":"AA-AnalystAgent","row_ids":["809f69850cfb92e35fdef7e6fcd0b405"],"score":17.39}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","peer_count":82,"percentile":24.69,"percentile_exact":{"denominator":81,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":78,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","peer_count":24,"percentile":17.39,"percentile_exact":{"denominator":23,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2}],"score":28.59,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":900},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":900},"family":"FrontierCode","row_ids":["eae24b4f084dd40f578e44715aebd484"],"score":18.37}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","peer_count":50,"percentile":18.37,"percentile_exact":{"denominator":49,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818}],"score":18.37,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":137425},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1625},"family":"GDP.pdf","row_ids":["0849343c6ca0b1de586bda128e33678b"],"score":42.76},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1350},"family":"AA-Omniscience","row_ids":["ce91d9d68242cca613b5b932f9a344d4","934c0d476e0913dedfc11ab20974d89e"],"score":25.47}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548}],"score":34.12,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":450},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":450},"family":"MMMU-Pro","row_ids":["681af2b5c57d4ca46641c38951c3d479"],"score":26.47}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069}],"score":26.47,"weight":10}],"eligible":true,"id":"m-bf7f21a45c57e0f7","in_window":true,"name":"Muse Glimmer","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":87,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ff2b5baedd439ff80001794125500940","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":49.15,"percentile_exact":{"denominator":59,"numerator":2900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.835353535353535,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"muse-glimmer:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.835353535353535},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f04dc7e5afab13c2812e16f8a7f09157","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"muse-glimmer:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"67c349c9897445ce219d7f8acc0e4dbe","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"Diamond","peer_count":64,"percentile":26.98,"percentile_exact":{"denominator":63,"numerator":1700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.091,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"muse-glimmer:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.091},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"142184a428fc22f5f377c4fcd96f6a10","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2 Hard","peer_count":52,"percentile":23.53,"percentile_exact":{"denominator":17,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.07,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"muse-glimmer:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4a499df1f84b1eaf0a172cfe97a95f4c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":57.69,"percentile_exact":{"denominator":13,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.235051546391753,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"muse-glimmer:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.235051546391753},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cab0e8633e6425575d15983329eb67d3","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v2.1","peer_count":82,"percentile":24.69,"percentile_exact":{"denominator":81,"numerator":2000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":792.29,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"muse-glimmer:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.29},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af8dab2781940ab31c2449f02ac9e502","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"muse-glimmer:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"af88768655cfd944c0cc2a22a6f5661c","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":26.32,"percentile_exact":{"denominator":19,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06831776269647931,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"muse-glimmer:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06831776269647931},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6b81d5a0603a102ca44c8276f19acede","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":78,"percentile":14.29,"percentile_exact":{"denominator":7,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":477.83,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"muse-glimmer:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":477.83},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6751eebce76aacc484338eb2b0237376","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA run","peer_count":19,"percentile":27.78,"percentile_exact":{"denominator":9,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.34735899731423453,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"muse-glimmer:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.34735899731423453},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"809f69850cfb92e35fdef7e6fcd0b405","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA pass^5","peer_count":24,"percentile":17.39,"percentile_exact":{"denominator":23,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.2,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"muse-glimmer:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"eae24b4f084dd40f578e44715aebd484","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"2.0","peer_count":50,"percentile":18.37,"percentile_exact":{"denominator":49,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0147431818181818,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"muse-glimmer:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0147431818181818},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0849343c6ca0b1de586bda128e33678b","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current AA protocol","peer_count":77,"percentile":42.76,"percentile_exact":{"denominator":38,"numerator":1625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.1,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"muse-glimmer:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ce91d9d68242cca613b5b932f9a344d4","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":39.62,"percentile_exact":{"denominator":53,"numerator":2100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.2796666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.2796666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Muse Glimmer (High)","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"934c0d476e0913dedfc11ab20974d89e","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.1642757982415548,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"muse-glimmer:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1642757982415548},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Muse Glimmer (High)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"681af2b5c57d4ca46641c38951c3d479","included":true,"model_id":"m-bf7f21a45c57e0f7","original_version_label":"current Standard 10","peer_count":35,"percentile":26.47,"percentile_exact":{"denominator":17,"numerator":450},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.743352601156069,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"muse-glimmer:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"muse-glimmer:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743352601156069}],"release_date":"2026-08-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":27.6,"score_exact":{"denominator":19177971701616,"numerator":529374170552879},"source_id":"meta/muse-glimmer-30b"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":63,"numerator":1000},"families":[{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1000},"family":"HLE","row_ids":["11ffdf294b4e40190b29f471a44eb232"],"score":15.87}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","peer_count":64,"percentile":15.87,"percentile_exact":{"denominator":63,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":15.87,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":248026779,"numerator":8390522060},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"APEX-Agents","row_ids":["ca1ee026c3875e1f8a6f67a549e74fe8"],"score":25.0},{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":1000},"family":"tau3-Banking","row_ids":["c993b26e42ea6045232f1d9e621ef2fa"],"score":76.92},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":3500},"family":"GDPval-AA","row_ids":["e73dd8b4c1ab25719febf181d0bca142"],"score":43.21},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["d43f0f17f32dc60cbed0bdb79fe20eb1"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":525},"family":"AutomationBench","row_ids":["7fbcd0afe5186232a5dba432e082b6ef"],"score":27.63},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2900},"family":"AA-Briefcase","row_ids":["11eea946a9e871701b9adaec2c3d14b2"],"score":37.66},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":300},"family":"AA-AnalystAgent","row_ids":["c283db79a5fa78b491d2c36ac505fd73"],"score":13.04}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","peer_count":82,"percentile":43.21,"percentile_exact":{"denominator":81,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":27.63,"percentile_exact":{"denominator":19,"numerator":525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","peer_count":78,"percentile":37.66,"percentile_exact":{"denominator":77,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","peer_count":24,"percentile":13.04,"percentile_exact":{"denominator":23,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625}],"score":33.83,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":1075},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":1075},"family":"GDP.pdf","row_ids":["0e26cf9d2beafb47ed1ed89556040c14"],"score":28.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"score":28.29,"weight":20}],"eligible":true,"id":"m-95ce1c534519e66a","in_window":true,"name":"Ling-3.0-flash-Fin","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":88,"records":[{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"11ffdf294b4e40190b29f471a44eb232","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"Diamond","peer_count":64,"percentile":15.87,"percentile_exact":{"denominator":63,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.054,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"ling-3-0-flash-fin:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ca1ee026c3875e1f8a6f67a549e74fe8","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA run","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; apex-agents-aa","protocol_id":"c455abb49f64fc5efd099eb8","raw_value":0.274336283185841,"reference_revision":"8d2dd74496317fbca6213e75ae0ae03870c2c5aeac56b40da7f86a1a96d3af9e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:apex-agents-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:APEX-Agents","source_record_id":"ling-3-0-flash-fin:apex-agents-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.274336283185841},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c993b26e42ea6045232f1d9e621ef2fa","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":76.92,"percentile_exact":{"denominator":13,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.385567010309278,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-flash-fin:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.385567010309278},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e73dd8b4c1ab25719febf181d0bca142","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v2.1","peer_count":82,"percentile":43.21,"percentile_exact":{"denominator":81,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1107.96,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-flash-fin:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1107.96},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d43f0f17f32dc60cbed0bdb79fe20eb1","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-flash-fin:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7fbcd0afe5186232a5dba432e082b6ef","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":27.63,"percentile_exact":{"denominator":19,"numerator":525},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.07394444993005693,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-flash-fin:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.07394444993005693},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"11eea946a9e871701b9adaec2c3d14b2","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"v1.1","peer_count":78,"percentile":37.66,"percentile_exact":{"denominator":77,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":958.63,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-flash-fin:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":958.63},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c283db79a5fa78b491d2c36ac505fd73","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA pass^5","peer_count":24,"percentile":13.04,"percentile_exact":{"denominator":23,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.1625,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"ling-3-0-flash-fin:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1625},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling-3.0-flash-Fin","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0e26cf9d2beafb47ed1ed89556040c14","included":true,"model_id":"m-95ce1c534519e66a","original_version_label":"current AA protocol","peer_count":77,"percentile":28.29,"percentile_exact":{"denominator":38,"numerator":1075},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.054,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-flash-fin:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-flash-fin:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.054}],"release_date":"2026-09-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":25.83,"score_exact":{"denominator":1736187453,"numerator":44851731725},"source_id":"inclusionai/ling-3.0-flash-fin"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":39117,"numerator":787925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1850},"family":"GPQA Diamond","row_ids":["885831acf33b9d6286fb559cc7b4ec0e"],"score":31.36},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d166f2e488fa71112bb60e275a7535f1"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["3441fc8b1a39080bc19e5b920fef2c87"],"score":11.76}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":20.14,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1100385,"numerator":49356484},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":800},"family":"tau3-Banking","row_ids":["33c50c79dd62f0151f2db3df92a057bc"],"score":61.54},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2900},"family":"GDPval-AA","row_ids":["4aa9f9d8e559b9430ab1f871a4ca4271"],"score":35.8},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":42},"family":"Terminal-Bench","row_ids":["5a757d2c757812d80a306fe7007417cd"],"score":42.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":750},"family":"AutomationBench","row_ids":["b87ec7052a30aebc07b1b22c9deef3cd"],"score":39.47},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":500},"family":"AA-Briefcase","row_ids":["6d3a1ea10b80de3baf1922e2ef5fc22e"],"score":45.45}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","peer_count":82,"percentile":35.8,"percentile_exact":{"denominator":81,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","peer_count":78,"percentile":45.45,"percentile_exact":{"denominator":11,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51}],"score":44.85,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":800},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":800},"family":"FrontierCode","row_ids":["d64bab068e454f76d91e1fb18d73dd68"],"score":16.33}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","peer_count":50,"percentile":16.33,"percentile_exact":{"denominator":49,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525}],"score":16.33,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":625},"family":"GDP.pdf","row_ids":["48f1c6876fabfb820ee570e743c9611a"],"score":16.45}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"score":16.45,"weight":20}],"eligible":true,"id":"m-c58d5bdb4d2e38bc","in_window":true,"name":"K2 Horizon 7B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":89,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"885831acf33b9d6286fb559cc7b4ec0e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":31.36,"percentile_exact":{"denominator":59,"numerator":1850},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.752525252525252,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.752525252525252},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d166f2e488fa71112bb60e275a7535f1","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3441fc8b1a39080bc19e5b920fef2c87","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"33c50c79dd62f0151f2db3df92a057bc","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":61.54,"percentile_exact":{"denominator":13,"numerator":800},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.276288659793814,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.276288659793814},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4aa9f9d8e559b9430ab1f871a4ca4271","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v2.1","peer_count":82,"percentile":35.8,"percentile_exact":{"denominator":81,"numerator":2900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":1024.31,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1024.31},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5a757d2c757812d80a306fe7007417cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"4.0","peer_count":76,"percentile":42.0,"percentile_exact":{"denominator":1,"numerator":42},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.0101010101010101,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0101010101010101},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b87ec7052a30aebc07b1b22c9deef3cd","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":39.47,"percentile_exact":{"denominator":19,"numerator":750},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.209005917720684,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.209005917720684},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"6d3a1ea10b80de3baf1922e2ef5fc22e","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"v1.1","peer_count":78,"percentile":45.45,"percentile_exact":{"denominator":11,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":1013.51,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":1013.51},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"d64bab068e454f76d91e1fb18d73dd68","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"2.0","peer_count":50,"percentile":16.33,"percentile_exact":{"denominator":49,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0116252525252525,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0116252525252525},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"48f1c6876fabfb820ee570e743c9611a","included":true,"model_id":"m-c58d5bdb4d2e38bc","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":25.34,"score_exact":{"denominator":194690237742,"numerator":4933003274473},"source_id":"published/aa:k2-horizon-7b"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":83020675},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2650},"family":"GPQA Diamond","row_ids":["c948234960090e8c9ef801d4c2e468fc"],"score":44.92},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["5d3cac863b9cb777dcf0429fca5619c7"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1150},"family":"HLE","row_ids":["1e3489d548cb804c3cfe4ff52df120e8"],"score":18.25},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":350},"family":"CritPt","row_ids":["55904790fd4c7a2e974218f474b3d773"],"score":20.59}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":44.92,"percentile_exact":{"denominator":59,"numerator":2650},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04}],"score":25.27,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":73359,"numerator":1386274},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":225},"family":"tau3-Banking","row_ids":["77cc539d6e52970112947681c17a7cf5"],"score":17.31},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":700},"family":"GDPval-AA","row_ids":["3eb5f8707daab0412d08dc253ada641e"],"score":25.93},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["017b50b4eb6f553388684608c12adbeb"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":375},"family":"AutomationBench","row_ids":["cf48341c74514d20a48b2f428df05755"],"score":19.74},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":200},"family":"AA-Briefcase","row_ids":["42d65fa891e8c3a4f35b4ea031798999"],"score":18.18}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","peer_count":82,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":78,"percentile":18.18,"percentile_exact":{"denominator":11,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09}],"score":18.9,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":500},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":500},"family":"FrontierCode","row_ids":["a86b3f52322966e22c9f2052d8fbc893"],"score":10.2}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","peer_count":50,"percentile":10.2,"percentile_exact":{"denominator":49,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404}],"score":10.2,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":4028,"numerator":119725},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":825},"family":"GDP.pdf","row_ids":["7e1c74ae3bf26c2c5b56bd7d43f2a5df"],"score":21.71},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2000},"family":"AA-Omniscience","row_ids":["6b5784505206d57d0bba2ceb5b8de557","3bcf44016a4447d1427e1bf692c1e1bc"],"score":37.74}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963}],"score":29.72,"weight":20}],"eligible":true,"id":"m-c46d8f7f1c3144cb","in_window":true,"name":"K-EXAONE 2.0 0803","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-12","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":90,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c948234960090e8c9ef801d4c2e468fc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":44.92,"percentile_exact":{"denominator":59,"numerator":2650},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.829292929292929,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"jt-35b-flash:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.829292929292929},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5d3cac863b9cb777dcf0429fca5619c7","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1e3489d548cb804c3cfe4ff52df120e8","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"55904790fd4c7a2e974218f474b3d773","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2 Hard","peer_count":52,"percentile":20.59,"percentile_exact":{"denominator":17,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.04,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"77cc539d6e52970112947681c17a7cf5","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.115463917525773,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-fable-5-1-low:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"claude-fable-5-1-low:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.115463917525773},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3eb5f8707daab0412d08dc253ada641e","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v2.1","peer_count":82,"percentile":25.93,"percentile_exact":{"denominator":27,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":815.8,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"claude-sonnet-5-5-medium:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":815.8},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"017b50b4eb6f553388684608c12adbeb","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"claude-sonnet-5-5-medium:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"cf48341c74514d20a48b2f428df05755","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":19.74,"percentile_exact":{"denominator":19,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.038837365237439365,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"claude-sonnet-5-5-medium:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.038837365237439365},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"42d65fa891e8c3a4f35b4ea031798999","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":78,"percentile":18.18,"percentile_exact":{"denominator":11,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":651.09,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"claude-sonnet-5-5-medium:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":651.09},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"a86b3f52322966e22c9f2052d8fbc893","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"2.0","peer_count":50,"percentile":10.2,"percentile_exact":{"denominator":49,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00831540404040404,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00831540404040404},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7e1c74ae3bf26c2c5b56bd7d43f2a5df","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"current AA protocol","peer_count":77,"percentile":21.71,"percentile_exact":{"denominator":38,"numerator":825},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.042,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"claude-sonnet-5-5-medium:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6b5784505206d57d0bba2ceb5b8de557","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":9.43,"percentile_exact":{"denominator":53,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.145,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.145},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"K-EXAONE 2.0","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3bcf44016a4447d1427e1bf692c1e1bc","included":true,"model_id":"m-c46d8f7f1c3144cb","original_version_label":"v1.1","peer_count":54,"percentile":66.04,"percentile_exact":{"denominator":53,"numerator":3500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6962962962962963,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6962962962962963}],"release_date":"2026-08-12","release_status":"published","release_verified":true,"result_updated_at":null,"score":21.14,"score_exact":{"denominator":13758110133768,"numerator":290851131141355},"source_id":"published/aa:k-exaone-2-0-0803"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":1500},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1500},"family":"GPQA Diamond","row_ids":["3e2102f0b8440ab308d1e2fedeb7c1ef"],"score":25.42}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":25.42,"percentile_exact":{"denominator":59,"numerator":1500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434}],"score":25.42,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3081078,"numerator":55952933},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":26,"numerator":1225},"family":"tau3-Banking","row_ids":["e330721d27c5f1f48579252fe1951f5e"],"score":47.12},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1300},"family":"GDPval-AA","row_ids":["f9e9baed830cb24bcff4431e5893ddf5"],"score":16.05},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["9fef6d5b9c29a36ef4ec13143b2d54ae"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":25},"family":"AutomationBench","row_ids":["a389be9490eced3c5e8431f3790a6ecd"],"score":1.32},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1000},"family":"AA-Briefcase","row_ids":["662e05367e55e07efb079daf8cc94493"],"score":12.99}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","peer_count":82,"percentile":16.05,"percentile_exact":{"denominator":81,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":78,"percentile":12.99,"percentile_exact":{"denominator":77,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91}],"score":18.16,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":75025},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":125},"family":"GDP.pdf","row_ids":["7738f3afb94846ed2b13073b7d2cb3b1"],"score":3.29},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["97a4da3275fb19d0e062184a2de4a918","ff2df3ac7046d17ae8e64e925f03f746"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413}],"score":18.63,"weight":20}],"eligible":true,"id":"m-6b5176ed51518e35","in_window":true,"name":"Ling 3.0 Tiny","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-06","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":91,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"3e2102f0b8440ab308d1e2fedeb7c1ef","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":25.42,"percentile_exact":{"denominator":59,"numerator":1500},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.734343434343434,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"ling-3-0-tiny:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.734343434343434},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e330721d27c5f1f48579252fe1951f5e","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":47.12,"percentile_exact":{"denominator":26,"numerator":1225},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.208247422680412,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"ling-3-0-tiny:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.208247422680412},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f9e9baed830cb24bcff4431e5893ddf5","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v2.1","peer_count":82,"percentile":16.05,"percentile_exact":{"denominator":81,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":568.16,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"ling-3-0-tiny:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":568.16},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"9fef6d5b9c29a36ef4ec13143b2d54ae","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"ling-3-0-tiny:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a389be9490eced3c5e8431f3790a6ecd","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002807334937855976,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"ling-3-0-tiny:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002807334937855976},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"662e05367e55e07efb079daf8cc94493","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":78,"percentile":12.99,"percentile_exact":{"denominator":77,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":475.91,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"ling-3-0-tiny:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":475.91},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"7738f3afb94846ed2b13073b7d2cb3b1","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"ling-3-0-tiny:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"97a4da3275fb19d0e062184a2de4a918","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.37933333333333336,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.37933333333333336},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Ling 3.0 Tiny","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"ff2df3ac7046d17ae8e64e925f03f746","included":true,"model_id":"m-6b5176ed51518e35","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.13319011815252413,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"ling-3-0-tiny:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"ling-3-0-tiny:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.13319011815252413}],"release_date":"2026-08-06","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.89,"score_exact":{"denominator":134883432684,"numerator":2817358651405},"source_id":"published/aa:ling-3-0-tiny"},{"benchmark_count":16,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":49583375},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1700},"family":"GPQA Diamond","row_ids":["f6a22fbde90853e88aebc327f5973f62"],"score":28.81},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["1b904acff7083099b3a91e7b6c5b575a"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":650},"family":"HLE","row_ids":["5c0267c76de42807699083ce1e0581c4"],"score":10.32},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":200},"family":"CritPt","row_ids":["98267d8baf13117a13191a66537ee6eb"],"score":3.92}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":28.81,"percentile_exact":{"denominator":59,"numerator":1700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","peer_count":52,"percentile":3.92,"percentile_exact":{"denominator":51,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005}],"score":15.09,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1450449,"numerator":28030885},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":350},"family":"tau3-Banking","row_ids":["d0f6a7a8e824abf26de5a2c3eb1a8991"],"score":26.92},{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":200},"family":"GDPval-AA","row_ids":["641dcee193d24a80d7b9a09327dfcf69"],"score":22.22},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["47e9d9d6a84d129959dab9eed70b71bc"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"AutomationBench","row_ids":["b05e70f346089460a993097793ed0c46"],"score":25.0},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1300},"family":"AA-Briefcase","row_ids":["671c80eaf4d486bab2d265a1c67045fd"],"score":16.88},{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"EnterpriseOps-Gym","row_ids":["da0d77fdb30d1e419d687691f9d0af7b"],"score":22.22},{"benchmark_version":"current AA pass^5","exact":{"denominator":23,"numerator":200},"family":"AA-AnalystAgent","row_ids":["ef6364df2840f04f7c6f211bd995980d"],"score":8.7}],"family_count":7,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","peer_count":82,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":78,"percentile":16.88,"percentile_exact":{"denominator":77,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","peer_count":24,"percentile":8.7,"percentile_exact":{"denominator":23,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125}],"score":19.33,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":1000},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":1000},"family":"FrontierCode","row_ids":["c2dd8ca14021780f9edc8dc0975ff8ae"],"score":20.41}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","peer_count":50,"percentile":20.41,"percentile_exact":{"denominator":49,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939}],"score":20.41,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":106,"numerator":3825},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":75},"family":"IFBench","row_ids":["5fbfa4015dcb4ab37c1080316055d055"],"score":75.0},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"GDP.pdf","row_ids":["2bd6dbb1c5200680f554dfce1a26feb0"],"score":12.5},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1100},"family":"AA-Omniscience","row_ids":["2b5b8d191f3e5e3bfa651cee2bbbd819","d7d18670af809e53c710911ecd277233"],"score":20.75}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378}],"score":36.08,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":150},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":150},"family":"MMMU-Pro","row_ids":["ce0faae345d00b20396f1024d947c26c"],"score":8.82}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798}],"score":8.82,"weight":10}],"eligible":true,"id":"m-5c18cc0e1d3391a9","in_window":true,"name":"Mistral Medium 3.5","official_evidence":{"artifact_sha256":"2f83ab54cf0df8335b080c25c95cc6852ada259bf85c12135f848a991e748f85","basis":"official_release_page","date":"2026-05-22","sha256":"2f83ab54cf0df8335b080c25c95cc6852ada259bf85c12135f848a991e748f85","url":"https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5/"},"primary_profile_id":"native","profile_id":"native","rank":92,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"f6a22fbde90853e88aebc327f5973f62","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":28.81,"percentile_exact":{"denominator":59,"numerator":1700},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.748484848484849,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"mistral-medium-3-5:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.748484848484849},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1b904acff7083099b3a91e7b6c5b575a","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"5c0267c76de42807699083ce1e0581c4","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"Diamond","peer_count":64,"percentile":10.32,"percentile_exact":{"denominator":63,"numerator":650},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.047,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mistral-medium-3-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.047},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"98267d8baf13117a13191a66537ee6eb","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2 Hard","peer_count":52,"percentile":3.92,"percentile_exact":{"denominator":51,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.005,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.005},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d0f6a7a8e824abf26de5a2c3eb1a8991","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":26.92,"percentile_exact":{"denominator":13,"numerator":350},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.150515463917526,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"mistral-medium-3-5:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.150515463917526},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"641dcee193d24a80d7b9a09327dfcf69","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v2.1","peer_count":82,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":768.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mistral-medium-3-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":768.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47e9d9d6a84d129959dab9eed70b71bc","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mistral-medium-3-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b05e70f346089460a993097793ed0c46","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.06299648656993127,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mistral-medium-3-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.06299648656993127},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"671c80eaf4d486bab2d265a1c67045fd","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":78,"percentile":16.88,"percentile_exact":{"denominator":77,"numerator":1300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":523.26,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mistral-medium-3-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":523.26},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"da0d77fdb30d1e419d687691f9d0af7b","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA run","peer_count":19,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.3372127723067741,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"mistral-medium-3-5:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.3372127723067741},{"benchmark_family":"AA-AnalystAgent","benchmark_id":"b-cf1d47e9f3bca5b555d0","benchmark_label":"AA-AnalystAgent","benchmark_version":"current AA pass^5","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ef6364df2840f04f7c6f211bd995980d","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA pass^5","peer_count":24,"percentile":8.7,"percentile_exact":{"denominator":23,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-analyst-agent","protocol_id":"0fdc20dc341754f0aa205de6","raw_value":0.125,"reference_revision":"6d84a5d8c0abe2d63236e646d24a418fabd31583e064f1e028ae563ea042bb03","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:aa-analyst-agent"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-AnalystAgent","source_record_id":"mistral-medium-3-5:aa-analyst-agent","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.125},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c2dd8ca14021780f9edc8dc0975ff8ae","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"2.0","peer_count":50,"percentile":20.41,"percentile_exact":{"denominator":49,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0176893939393939,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0176893939393939},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5fbfa4015dcb4ab37c1080316055d055","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current 294-item protocol","peer_count":17,"percentile":75.0,"percentile_exact":{"denominator":1,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.687755102040816,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"mistral-medium-3-5:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.687755102040816},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2bd6dbb1c5200680f554dfce1a26feb0","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mistral-medium-3-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2b5b8d191f3e5e3bfa651cee2bbbd819","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":26.42,"percentile_exact":{"denominator":53,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.23033333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.23033333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Mistral Medium 3.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d7d18670af809e53c710911ecd277233","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"v1.1","peer_count":54,"percentile":15.09,"percentile_exact":{"denominator":53,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.19792117799913378,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"mistral-medium-3-5:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.19792117799913378},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mistral Medium 3.5","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"ce0faae345d00b20396f1024d947c26c","included":true,"model_id":"m-5c18cc0e1d3391a9","original_version_label":"current Standard 10","peer_count":35,"percentile":8.82,"percentile_exact":{"denominator":17,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.648554913294798,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mistral-medium-3-5:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"mistral-medium-3-5:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.648554913294798}],"release_date":"2026-05-22","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.78,"score_exact":{"denominator":411223564752,"numerator":8547240422815},"source_id":"mistral/mistral-medium-3-5"},{"benchmark_count":5,"configuration":"公开协议汇总","coverage":2,"coverage_percent":40.0,"dimensions":[{"effective_weight_percent":55.56,"evaluator_count":1,"exact":{"denominator":474012,"numerator":12975565},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2600},"family":"GDPval-AA","row_ids":["a6f95ea45eef0a446f5d6ed10bf3ca1b"],"score":32.1},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["614dafc454b6d64e61b2f25573ec4717"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":625},"family":"AutomationBench","row_ids":["d9c233b856dc8ecc623629216c3b1c31"],"score":32.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":2400},"family":"AA-Briefcase","row_ids":["e9c29563b5cc25446296aa38a0ecd4ae"],"score":31.17}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","peer_count":82,"percentile":32.1,"percentile_exact":{"denominator":81,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","peer_count":78,"percentile":31.17,"percentile_exact":{"denominator":77,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94}],"score":27.37,"weight":25},{"effective_weight_percent":44.44,"evaluator_count":1,"exact":{"denominator":2,"numerator":25},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"GDP.pdf","row_ids":["b6eecb655a01ff6addbb4005e50d5c51"],"score":12.5}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028}],"score":12.5,"weight":20}],"eligible":true,"id":"m-3ad11f2b6253b364","in_window":true,"name":"Nemotron 3.5 Bolt 74B-A9B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-10-15","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":93,"records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a6f95ea45eef0a446f5d6ed10bf3ca1b","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v2.1","peer_count":82,"percentile":32.1,"percentile_exact":{"denominator":81,"numerator":2600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":979.03,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":979.03},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"614dafc454b6d64e61b2f25573ec4717","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-bolt-74b-a9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d9c233b856dc8ecc623629216c3b1c31","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":32.89,"percentile_exact":{"denominator":19,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.1546088064040565,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-bolt-74b-a9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.1546088064040565},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e9c29563b5cc25446296aa38a0ecd4ae","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"v1.1","peer_count":78,"percentile":31.17,"percentile_exact":{"denominator":77,"numerator":2400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":911.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-bolt-74b-a9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":911.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Bolt 74B-A9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b6eecb655a01ff6addbb4005e50d5c51","included":true,"model_id":"m-3ad11f2b6253b364","original_version_label":"current AA protocol","peer_count":77,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.028,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-bolt-74b-a9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-bolt-74b-a9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.028}],"release_date":"2026-10-15","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.76,"score_exact":{"denominator":4266108,"numerator":88578425},"source_id":"published/aa:nemotron-3-5-bolt-74b-a9b"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":1000},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1000},"family":"GPQA Diamond","row_ids":["9804205b8a1d620728568c184381a38c"],"score":16.95}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":16.95,"percentile_exact":{"denominator":59,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869}],"score":16.95,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":27,"numerator":200},"families":[{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":200},"family":"GDPval-AA","row_ids":["4fa0d328e940e0ba1d47d7e70c6e0521"],"score":7.41}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","peer_count":82,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5}],"score":7.41,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":4,"numerator":175},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":175},"family":"IFBench","row_ids":["8e2367487c04d7937c7f24678241f073"],"score":43.75}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252}],"score":43.75,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":17,"numerator":200},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":200},"family":"MMMU-Pro","row_ids":["d47a048ab5b9e3796a7da71c7471a272"],"score":11.76}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144}],"score":11.76,"weight":10}],"eligible":true,"id":"m-fe0a26f770e1e011","in_window":true,"name":"DiffusionGemma 26B A4B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-10","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":94,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9804205b8a1d620728568c184381a38c","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":16.95,"percentile_exact":{"denominator":59,"numerator":1000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.668686868686869,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"diffusiongemma-26b-a4b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.668686868686869},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"4fa0d328e940e0ba1d47d7e70c6e0521","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"v2.1","peer_count":82,"percentile":7.41,"percentile_exact":{"denominator":27,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":328.5,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"diffusiongemma-26b-a4b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":328.5},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8e2367487c04d7937c7f24678241f073","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current 294-item protocol","peer_count":17,"percentile":43.75,"percentile_exact":{"denominator":4,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.594557823129252,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"diffusiongemma-26b-a4b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.594557823129252},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"DiffusionGemma 26B A4B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"d47a048ab5b9e3796a7da71c7471a272","included":true,"model_id":"m-fe0a26f770e1e011","original_version_label":"current Standard 10","peer_count":35,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.665317919075144,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"diffusiongemma-26b-a4b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"diffusiongemma-26b-a4b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.665317919075144}],"release_date":"2026-06-10","release_status":"published","release_verified":true,"result_updated_at":null,"score":20.02,"score_exact":{"denominator":433296,"numerator":8674375},"source_id":"published/aa:diffusiongemma-26b-a4b"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":19,"numerator":200},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":0},"family":"FrontierMath","row_ids":["040f48951e0ebb9ff6bfbde83c816f48"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":1,"numerator":0},"family":"GPQA Diamond","row_ids":["29bbe2be4ccedeb019b3ccf6a3404eb9"],"score":0.0},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":600},"family":"Terminal-Bench-Science","row_ids":["540adb8175ab28d08c531a963e348646"],"score":31.58}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v2","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29}],"score":10.53,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":5,"exact":{"denominator":3,"numerator":124},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":40},"family":"APEX-Agents","row_ids":["9584a53c71e055a6a275ffd8975b0b25"],"score":40.0},{"benchmark_version":"4.0","exact":{"denominator":6,"numerator":275},"family":"Terminal-Bench","row_ids":["2539da8c839981606943a4efe846b325"],"score":45.83},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"AutomationBench","row_ids":["6bd66300b839f4ff34d5fd8249350449"],"score":37.5},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":50},"family":"AA-Briefcase","row_ids":["52ae106aa8230eabcaac4637b6d547ea"],"score":50.0},{"benchmark_version":"1000 tasks","exact":{"denominator":3,"numerator":100},"family":"MCP Atlas","row_ids":["4cec2c81b7ca1a60e4ad36412d5896a7"],"score":33.33}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","peer_count":49,"percentile":45.83,"percentile_exact":{"denominator":6,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2}],"score":41.33,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["dee93fd91e74e8ba6728dba7ca4fd982"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0}],"score":0.0,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":3,"exact":{"denominator":459,"numerator":11950},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":200},"family":"IFBench","row_ids":["0a5ea9e2a9d5f26a6a034bb3853216a5"],"score":22.22},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"GDP.pdf","row_ids":["dc77f0ff3c077230aef1b2b8cb04ec92"],"score":0.0},{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":950},"family":"SimpleQA Verified","row_ids":["dbd9d582709dbd41201aa5f60842dfb0"],"score":55.88}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53}],"score":26.03,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":9,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":9,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["2c5710356bbeeb960760bc793dbd6d80"],"score":11.11},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["24b2d438f8c9ef458157a109e7bcb701"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","peer_count":28,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9}],"score":5.56,"weight":10}],"eligible":true,"id":"m-cccbc1ce3042d216","in_window":true,"name":"Claude Opus 4.8","official_evidence":{"artifact_sha256":"f43b616507c976c560385970d34c456a9634c2e81776613c8670c64ab10e5e35","basis":"official_release_page","date":"2026-05-28","sha256":"f43b616507c976c560385970d34c456a9634c2e81776613c8670c64ab10e5e35","url":"https://www.anthropic.com/news/claude-opus-4-8"},"primary_profile_id":"native","profile_id":"native","rank":95,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"040f48951e0ebb9ff6bfbde83c816f48","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v2","peer_count":8,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 4 (v2); preserve page footnotes and task-specific settings","protocol_id":"8d02db73f46a4d9d6e92f159","raw_value":56.1,"reference_revision":"c966d842f112e0d9c6ce525e2cff2a037a51d17a956db6f70d19564c09ffcc68","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row3:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 4 (v2)","source_record_id":"openai56:table7:row3:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":56.1},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"reasoning","direction":"higher_better","id":"29bbe2be4ccedeb019b3ccf6a3404eb9","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column Claude Opus 4.8. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.0,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.0},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"540adb8175ab28d08c531a963e348646","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"0.1.0","peer_count":39,"percentile":31.58,"percentile_exact":{"denominator":19,"numerator":600},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":4.29,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"4.29%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row30"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row30","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":4.29},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"9584a53c71e055a6a275ffd8975b0b25","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Current published protocol","peer_count":6,"percentile":40.0,"percentile_exact":{"denominator":1,"numerator":40},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column Claude Opus 4.8 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":39.4,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"39.4","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":39.4},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"Claude Opus 4.8 / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"2539da8c839981606943a4efe846b325","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"4.0","peer_count":49,"percentile":45.83,"percentile_exact":{"denominator":6,"numerator":275},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":23.23,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"23.23%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row29"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row29","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":23.23},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"agents","direction":"higher_better","id":"6bd66300b839f4ff34d5fd8249350449","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"published developer protocol","peer_count":9,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":15.5,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column8"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column8","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":15.5},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"Claude Opus 4.8 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"52ae106aa8230eabcaac4637b6d547ea","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"v1.1","peer_count":19,"percentile":50.0,"percentile_exact":{"denominator":1,"numerator":50},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.38686868686868686,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"eb7d9b043faaf96f99309e41"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"eb7d9b043faaf96f99309e41","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.38686868686868686},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-08T03:27:51.596147+00:00","component_key":"score","configuration":"claude-opus-4-8 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-scale-mcp","id":"4cec2c81b7ca1a60e4ad36412d5896a7","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1000 / current public protocol","peer_count":10,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"scale-mcp-reviewed-method-20261029","protocol_id":"89dd905180d24f13af9f3f8e","raw_value":82.2,"reference_revision":"6c9f7728769f1bd8323172ef2102caf8a73899dd01a925e86eb5002aad78978e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"f5eacc0779e8f39e6d645139ec10711ff84365ba868713841a2820da6d6d5b43","result_date":null,"source_cell":null,"source_id":"s-6e504849b92015a53acf","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d6b730c39df07248ec639642"},"source_name":"Scale AI","source_protocol_key":"scale-mcp-reviewed-method","source_record_id":"d6b730c39df07248ec639642","source_url":"https://labs.scale.com/leaderboard/mcp_atlas","unit":"%","value":82.2},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"coding","direction":"higher_better","id":"dee93fd91e74e8ba6728dba7ca4fd982","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"1.1","peer_count":9,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":59.0,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column9"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column9","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":59.0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:27:26.166396+00:00","component_key":"score","configuration":"Opus 4.8","dimension":"language_information","direction":"higher_better","feed_id":"public-release-2bc2cf30ffd56fe1","id":"0a5ea9e2a9d5f26a6a034bb3853216a5","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":10,"percentile":22.22,"percentile_exact":{"denominator":9,"numerator":200},"profile_key":"native","protocol_description":"Published page qwen38-weights; row IFBench; preserve page footnotes and task-specific settings","protocol_id":"81d4b10b8d99c4374542c3ff","raw_value":62.2,"reference_revision":"8e3023061e679bbda23d872d725c2f03f38ea5faf66d8f83f593f3e164793844","research_artifact_sha256":"cb68a9ec8a65f44375d215605ee195ffdfd3359235d296eca412ea1a03e52198","response_sha256":"f73ef71297ea103a78f388e94c4fe29d6905a36dcf929893d04aef9e15329905","result_date":null,"source_cell":"62.2","source_id":"s-cf62ad1f26e4ee4bb415","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"qwen38-weights:table0:row27:column1"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFBench:IFBench","source_record_id":"qwen38-weights:table0:row27:column1","source_url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","unit":"%","value":62.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"Claude Opus 4.8","dimension":"language_information","direction":"higher_better","id":"dc77f0ff3c077230aef1b2b8cb04ec92","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current published protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.5,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column6"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column6","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.5},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"claude-opus-4-8_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"dbd9d582709dbd41201aa5f60842dfb0","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"run 1.2.0","peer_count":35,"percentile":55.88,"percentile_exact":{"denominator":17,"numerator":950},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.53,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:27:24.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"904b698ed20b1dcf82c45f22"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"904b698ed20b1dcf82c45f22","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.53},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"Claude-Opus-4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"2c5710356bbeeb960760bc793dbd6d80","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"current public protocol","peer_count":28,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":23.5,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"46054b42173f00b6036e834b"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"46054b42173f00b6036e834b","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":23.5},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"Claude Opus 4.8 (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"24b2d438f8c9ef458157a109e7bcb701","included":true,"model_id":"m-cccbc1ce3042d216","original_version_label":"Standard 10","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row MMMU-Pro; preserve page footnotes and task-specific settings","protocol_id":"6fde3ba0bea9afec4846158a","raw_value":78.9,"reference_revision":"75b8b6334f610bb92755d323f6082df9b234f306b772dce489c6d181a7637ac1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"78.9 / 82.7","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row46:column4"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMMU-Pro:MMMU-Pro","source_record_id":"kimi-k3:table1:row46:column4","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":78.9}],"release_date":"2026-05-28","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":18.73,"score_exact":{"denominator":8721,"numerator":163322},"source_id":"anthropic/claude-opus-4-8"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":117351,"numerator":1674775},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1100},"family":"GPQA Diamond","row_ids":["c96afdd76392eb5d2c9b157b68b7de9e"],"score":18.64},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["d7eec79b902bf9a304fdfc120e9aa124"],"score":17.31},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":350},"family":"CritPt","row_ids":["9fef9d533791c4fe2ca276032af2b566"],"score":6.86}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":18.64,"percentile_exact":{"denominator":59,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01}],"score":14.27,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":42819149},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":550},"family":"tau3-Banking","row_ids":["7b39b47195e57aecded4286410a368f3"],"score":42.31},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":2300},"family":"GDPval-AA","row_ids":["eada0c68a6bf5686fb53166d5b815ba4"],"score":28.4},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["585ae90ea798ef43916073acc85d178e"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":575},"family":"AutomationBench","row_ids":["eb41669f3fb02b0d8fee80ec8c840979"],"score":30.26},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1900},"family":"AA-Briefcase","row_ids":["ed41f889fc4e7786b137bd6b76a3da51"],"score":24.68}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","peer_count":82,"percentile":28.4,"percentile_exact":{"denominator":81,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":30.26,"percentile_exact":{"denominator":19,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","peer_count":78,"percentile":24.68,"percentile_exact":{"denominator":77,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97}],"score":27.79,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":600},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":600},"family":"FrontierCode","row_ids":["908a5049e650c9f2841b87a654313ae7"],"score":12.24}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","peer_count":50,"percentile":12.24,"percentile_exact":{"denominator":49,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798}],"score":12.24,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":38,"numerator":625},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":625},"family":"GDP.pdf","row_ids":["e8733c3c5f23f9a2dee5ae8306fe34a6"],"score":16.45}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"score":16.45,"weight":20}],"eligible":true,"id":"m-26bf3900b3ceb955","in_window":true,"name":"K2 Horizon 3.7B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":96,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c96afdd76392eb5d2c9b157b68b7de9e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":18.64,"percentile_exact":{"denominator":59,"numerator":1100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.691919191919192,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-3-7b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.691919191919192},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d7eec79b902bf9a304fdfc120e9aa124","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"9fef9d533791c4fe2ca276032af2b566","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7b39b47195e57aecded4286410a368f3","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":42.31,"percentile_exact":{"denominator":13,"numerator":550},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.195876288659794,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-3-7b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.195876288659794},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eada0c68a6bf5686fb53166d5b815ba4","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v2.1","peer_count":82,"percentile":28.4,"percentile_exact":{"denominator":81,"numerator":2300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":888.82,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-3-7b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":888.82},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"585ae90ea798ef43916073acc85d178e","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-3-7b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"eb41669f3fb02b0d8fee80ec8c840979","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":30.26,"percentile_exact":{"denominator":19,"numerator":575},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.14695137139561937,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-3-7b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14695137139561937},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed41f889fc4e7786b137bd6b76a3da51","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"v1.1","peer_count":78,"percentile":24.68,"percentile_exact":{"denominator":77,"numerator":1900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":792.97,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-3-7b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":792.97},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"908a5049e650c9f2841b87a654313ae7","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"2.0","peer_count":50,"percentile":12.24,"percentile_exact":{"denominator":49,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.00905479797979798,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00905479797979798},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 3.7B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e8733c3c5f23f9a2dee5ae8306fe34a6","included":true,"model_id":"m-26bf3900b3ceb955","original_version_label":"current AA protocol","peer_count":77,"percentile":16.45,"percentile_exact":{"denominator":38,"numerator":625},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.032,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-3-7b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-3-7b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.032}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":18.06,"score_exact":{"denominator":97345118871,"numerator":1758167195585},"source_id":"published/aa:k2-horizon-3-7b"},{"benchmark_count":11,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":41628025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":900},"family":"GPQA Diamond","row_ids":["651c19da954966c3a71036c8d3db174d"],"score":15.25},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["47e73ec519c6695efcb0b041ef5795ce"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":400},"family":"HLE","row_ids":["db2ce17572462c1eb76d15ea72c97e2b"],"score":6.35},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["d5d41ba5c0f73bb7b22fdca4916cb1b7"],"score":11.76}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":15.25,"percentile_exact":{"denominator":59,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","peer_count":64,"percentile":6.35,"percentile_exact":{"denominator":63,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":12.67,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":118503,"numerator":1813883},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":25},"family":"tau3-Banking","row_ids":["767893d84c1938b3d363a56f8204eb3b"],"score":25.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1400},"family":"GDPval-AA","row_ids":["a655895858c8427a30e33e8793a2117d"],"score":17.28},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["027860a90cc2579f5f5af8c36e50c87e"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":200},"family":"AutomationBench","row_ids":["354c0d77cbcd6264333d316fdf9f676b"],"score":10.53},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":800},"family":"AA-Briefcase","row_ids":["afdcb1efb0aca22bea9c16a3d016167d"],"score":10.39}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","peer_count":82,"percentile":17.28,"percentile_exact":{"denominator":81,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":10.53,"percentile_exact":{"denominator":19,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":78,"percentile":10.39,"percentile_exact":{"denominator":77,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75}],"score":15.31,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":4028,"numerator":107275},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":375},"family":"GDP.pdf","row_ids":["6e95a209bd026922387753b507be18e1"],"score":9.87},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2300},"family":"AA-Omniscience","row_ids":["d19df36b3cd3a0e840c1256cd1c744fe","82eab72497730b1d30475a9ca4c7819b"],"score":43.4}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948}],"score":26.63,"weight":20}],"eligible":true,"id":"m-9453d55ff407aaee","in_window":true,"name":"Granite 4.2 30B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":97,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"651c19da954966c3a71036c8d3db174d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":15.25,"percentile_exact":{"denominator":59,"numerator":900},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.644444444444444,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-30b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.644444444444444},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"47e73ec519c6695efcb0b041ef5795ce","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"db2ce17572462c1eb76d15ea72c97e2b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"Diamond","peer_count":64,"percentile":6.35,"percentile_exact":{"denominator":63,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.043,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"claude-sonnet-5-5-medium:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.043},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"d5d41ba5c0f73bb7b22fdca4916cb1b7","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"767893d84c1938b3d363a56f8204eb3b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.144329896907216,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-30b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.144329896907216},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a655895858c8427a30e33e8793a2117d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v2.1","peer_count":82,"percentile":17.28,"percentile_exact":{"denominator":81,"numerator":1400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":582.87,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-30b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":582.87},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"027860a90cc2579f5f5af8c36e50c87e","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-30b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"354c0d77cbcd6264333d316fdf9f676b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":10.53,"percentile_exact":{"denominator":19,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.015132727566596104,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-30b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015132727566596104},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"afdcb1efb0aca22bea9c16a3d016167d","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":78,"percentile":10.39,"percentile_exact":{"denominator":77,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":394.75,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-30b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":394.75},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6e95a209bd026922387753b507be18e1","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-30b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-30b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d19df36b3cd3a0e840c1256cd1c744fe","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":3.77,"percentile_exact":{"denominator":53,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11233333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11233333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 30B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"82eab72497730b1d30475a9ca4c7819b","included":true,"model_id":"m-9453d55ff407aaee","original_version_label":"v1.1","peer_count":54,"percentile":83.02,"percentile_exact":{"denominator":53,"numerator":4400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.7333834021779948,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.7333834021779948}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":17.6,"score_exact":{"denominator":4586036711256,"numerator":80716639816145},"source_id":"published/aa:granite-4-2-30b"},{"benchmark_count":10,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":2,"exact":{"denominator":114,"numerator":625},"families":[{"benchmark_version":"2","exact":{"denominator":1,"numerator":0},"family":"FrontierMath","row_ids":["0c9a231c5ad369432b2d4e6968b6bc7c"],"score":0.0},{"benchmark_version":"Diamond","exact":{"denominator":2,"numerator":25},"family":"GPQA Diamond","row_ids":["710e7d489e0ed28e0e46334dd5f83f89"],"score":12.5},{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["85fbb43a0855c52959cd8ec4413297f0"],"score":3.95}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v2","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0}],"score":5.48,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":3,"exact":{"denominator":54,"numerator":1475},"families":[{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":25},"family":"Terminal-Bench","row_ids":["5bec3b19db8e7a3362bcd865766b2cc8"],"score":25.0},{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":25},"family":"AutomationBench","row_ids":["51dfce6bb3b25136b0ea3545031f03b8"],"score":12.5},{"benchmark_version":"1.1","exact":{"denominator":9,"numerator":400},"family":"AA-Briefcase","row_ids":["06516697c2ae14f64607af00c620c8a8"],"score":44.44}],"family_count":3,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","peer_count":49,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798}],"score":27.31,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":2,"numerator":25},"families":[{"benchmark_version":"1.1","exact":{"denominator":2,"numerator":25},"family":"DeepSWE","row_ids":["b632e7c137b19cc34931e9481ad27bf9"],"score":12.5}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2}],"score":12.5,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":25},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":25},"family":"GDP.pdf","row_ids":["706d4a87bdc7adf93c73d7e4099021c8"],"score":25.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7}],"score":25.0,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":2,"exact":{"denominator":27,"numerator":50},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":27,"numerator":100},"family":"Humanity Sixth Sense","row_ids":["6698824b06580d547cc03c11c24e9e05"],"score":3.7},{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["b27c843f9cde1b388c12022729f0ceb0"],"score":0.0}],"family_count":2,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","peer_count":28,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"reference_revision":"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4}],"score":1.85,"weight":10}],"eligible":true,"id":"m-65a8fc1dec61133f","in_window":true,"name":"GPT-5.6 Luna","official_evidence":{"artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","basis":"official_release_page","date":"2026-07-09","sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","url":"https://openai.com/index/gpt-5-6/"},"primary_profile_id":"native","profile_id":"native","rank":98,"records":[{"benchmark_family":"FrontierMath","benchmark_id":"b-1e92ed1ca952b7e5309d","benchmark_label":"FrontierMath","benchmark_version":"2","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"0c9a231c5ad369432b2d4e6968b6bc7c","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v2","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row FrontierMath Tier 1-3 (v2); preserve page footnotes and task-specific settings","protocol_id":"3842bb51683d99f4ac9c82c3","raw_value":78.6,"reference_revision":"b170a13a9fe892a591aa8af9da80f383e9baa2875d99704441cdd3f6fce88d6e","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row2:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:FrontierMath:FrontierMath Tier 1-3 (v2)","source_record_id":"openai56:table7:row2:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.6},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"reasoning","direction":"higher_better","id":"710e7d489e0ed28e0e46334dd5f83f89","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Diamond; fixed 198-question subset","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GPT‑5.6 Luna. Footnotes and all competing runs retained in source.","protocol_id":"032545e2298934b8fc66f83f","raw_value":92.3,"reference_revision":"80da499e6e0ca7332df2eb162b4ddc9c987c961127abe5958cde133f7a959508","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":"92.3%","source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table7:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GPQA Diamond:GPQA Diamond","source_record_id":"openai56:table7:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":92.3},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"85fbb43a0855c52959cd8ec4413297f0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row41"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row41","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:28:22.713088+00:00","component_key":"score","configuration":"GPT-5.6 Luna / Mini-SWE-agent","dimension":"agents","direction":"higher_better","feed_id":"public-vals-terminalbench","id":"5bec3b19db8e7a3362bcd865766b2cc8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"4.0","peer_count":49,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":11.62,"reference_revision":"aee548648bd88233a486905226e2fd175550aca425dedc1620ad26a2b5599a6d","research_artifact_sha256":"a6215981bc3ec7e98df0272da73f85c6479bbf7e47bc8a38344537acbec771c7","response_sha256":"aa603408f7a083e88869de516c1a8a43e54a06c37804794cccde5b9d2b31a27e","result_date":null,"source_cell":"11.62%","source_id":"s-549f084101bc19fe9031","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal4:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal4:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-4","unit":"%","value":11.62},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"agents","direction":"higher_better","id":"51dfce6bb3b25136b0ea3545031f03b8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"published developer protocol","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"a93de83dd8cfc372c687b91d","raw_value":14.9,"reference_revision":"b75a92f520886f7f263bd92b85eb0ed22deae253037834bd5e198cfaf9bc2454","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table8:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:AutomationBench:AutomationBench","source_record_id":"openai56:table8:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":14.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GPT-5.6 Luna (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"06516697c2ae14f64607af00c620c8a8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"v1.1","peer_count":19,"percentile":44.44,"percentile_exact":{"denominator":9,"numerator":400},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3797979797979798,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"9abc796b925f35214c3d86a1"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"9abc796b925f35214c3d86a1","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3797979797979798},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"coding","direction":"higher_better","id":"b632e7c137b19cc34931e9481ad27bf9","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"1.1","peer_count":9,"percentile":12.5,"percentile_exact":{"denominator":2,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row DeepSWE v1.1; preserve page footnotes and task-specific settings","protocol_id":"1b6f5b9c917154d5a53d4023","raw_value":67.2,"reference_revision":"2c3931d421543bd28d99b6a0ad5c73f02e20f2d66a5549998726e381502b041a","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table1:row3:column4"},"source_name":"openai.com","source_protocol_key":"official:openai.com:DeepSWE:DeepSWE v1.1","source_record_id":"openai56:table1:row3:column4","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":67.2},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"language_information","direction":"higher_better","id":"706d4a87bdc7adf93c73d7e4099021c8","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current published protocol","peer_count":5,"percentile":25.0,"percentile_exact":{"denominator":1,"numerator":25},"profile_key":"native","protocol_description":"Published page openai56; row gdp.pdf; preserve page footnotes and task-specific settings","protocol_id":"7ca6f7acd460b88b1d816a01","raw_value":22.7,"reference_revision":"399d1f37c00ed6822f95fb01f88479584f9f5b1b5e8c4c91ec9c14037f5a7ae7","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row3:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:GDP.pdf:gdp.pdf","source_record_id":"openai56:table6:row3:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":22.7},{"benchmark_family":"Humanity Sixth Sense","benchmark_id":"b-6cfe59f63f128c8c6304","benchmark_label":"Humanity Sixth Sense","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:30.674655+00:00","component_key":"score","configuration":"GPT-5.6-Luna (max)","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-release-dd11f0d368b627c2","id":"6698824b06580d547cc03c11c24e9e05","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"current public protocol","peer_count":28,"percentile":3.7,"percentile_exact":{"denominator":27,"numerator":100},"profile_key":"native","protocol_description":"scale-hss-reviewed-method-20261029","protocol_id":"e86182c6397db78c5990f7ac","raw_value":21.6,"reference_revision":"156c2526a070a17dea760ba9b175ea503b9cfe293ed188a17169eba38105895e","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"54f7c85a9ad9c2ea23b603aa6f8c5035aca0f5b1bfeab39711f6be9cf6f48a7f","result_date":null,"source_cell":null,"source_id":"s-fbaa49a206de7fcc239b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"d898985df3ef60e9db3fef57"},"source_name":"Scale AI","source_protocol_key":"scale-hss-reviewed-method","source_record_id":"d898985df3ef60e9db3fef57","source_url":"https://labs.scale.com/leaderboard/hss","unit":"%","value":21.6},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-10-29T10:00:59.694526+00:00","component_key":"score","configuration":"GPT‑5.6 Luna","dimension":"multimodal_understanding","direction":"higher_better","id":"b27c843f9cde1b388c12022729f0ceb0","included":true,"model_id":"m-65a8fc1dec61133f","original_version_label":"Standard 10","peer_count":3,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page openai56; row MMMU Pro (no tools); preserve page footnotes and task-specific settings","protocol_id":"c9f689bb3d88b096fb77bc89","raw_value":78.4,"reference_revision":"545da3ea5f1e3c4c974f0222aff3a426ccbe4ca9bc4fa08dba69c524e8f762ec","research_artifact_sha256":"58e19c15c017fac24448895ef8d9c7dda1b58423f3ddde5f50a5a352ce2f88d7","response_sha256":"8b7c3395e8c43a1cd7b2ac85ba68973b4498c2d6682a2c99482ce42f33d66ef6","result_date":"2026-07-09","source_cell":null,"source_id":"s-fb04d3e999c5866880bb","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"openai56:table6:row1:column3"},"source_name":"openai.com","source_protocol_key":"official:openai.com:MMMU-Pro:MMMU Pro (no tools)","source_record_id":"openai56:table6:row1:column3","source_url":"https://openai.com/index/gpt-5-6/","unit":"%","value":78.4}],"release_date":"2026-07-09","release_status":"published","release_verified":true,"result_updated_at":"2026-07-09","score":15.88,"score_exact":{"denominator":684,"numerator":10865},"source_id":"openai/gpt-5.6-luna"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":52678025},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":1600},"family":"GPQA Diamond","row_ids":["02699c8c436679b49b6319aa832d4d14"],"score":27.12},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["ac28090aae70acdb086bd22103c739d0"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":500},"family":"HLE","row_ids":["e12a9fbbd9fdb2bbd3147c3b6679efcd"],"score":7.94},{"benchmark_version":"2 Hard","exact":{"denominator":17,"numerator":200},"family":"CritPt","row_ids":["2f96f5af894d02c94d57177465def23b"],"score":11.76}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":27.12,"percentile_exact":{"denominator":59,"numerator":1600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond","peer_count":64,"percentile":7.94,"percentile_exact":{"denominator":63,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015}],"score":16.03,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":22949483},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":200},"family":"tau3-Banking","row_ids":["612f78abda8cf62d29f104d80c127636"],"score":15.38},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":500},"family":"GDPval-AA","row_ids":["1694dd426967f516c30dcdcc5616df9b"],"score":18.52},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["803b4920eccad03514634ebd3575a2d8"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":150},"family":"AutomationBench","row_ids":["ec3633dbe4490f5d8334e3da32d471af"],"score":7.89},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":1200},"family":"AA-Briefcase","row_ids":["47de14e8ef5807902d31f4e62879bab3"],"score":15.58},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"EnterpriseOps-Gym","row_ids":["80ae66f2b7393fe1f51bab7846dd0a10"],"score":0.0}],"family_count":6,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2.1","peer_count":82,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":78,"percentile":15.58,"percentile_exact":{"denominator":77,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA run","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537}],"score":14.9,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["c503c45f721e54502bd0c3c277c347cb"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":46875},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":275},"family":"GDP.pdf","row_ids":["40a272cc2b1879bb5b0104ebb8833d18"],"score":14.47},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1700},"family":"AA-Omniscience","row_ids":["8698ac189bdac0fe4a6655a8ed1ade73","2184077575074c69e38c71abae562f14"],"score":32.08}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636}],"score":23.27,"weight":20}],"eligible":true,"id":"m-9ff6170077842b96","in_window":true,"name":"Nemotron 3.5 Lightning","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":99,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"02699c8c436679b49b6319aa832d4d14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":27.12,"percentile_exact":{"denominator":59,"numerator":1600},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.743434343434343,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"nemotron-3-5-lightning:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.743434343434343},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"ac28090aae70acdb086bd22103c739d0","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"e12a9fbbd9fdb2bbd3147c3b6679efcd","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"Diamond","peer_count":64,"percentile":7.94,"percentile_exact":{"denominator":63,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.045,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"nemotron-3-5-lightning:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.045},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"2f96f5af894d02c94d57177465def23b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2 Hard","peer_count":52,"percentile":11.76,"percentile_exact":{"denominator":17,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.015,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.015},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"612f78abda8cf62d29f104d80c127636","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":15.38,"percentile_exact":{"denominator":13,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.088659793814433,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"nemotron-3-5-lightning:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.088659793814433},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"1694dd426967f516c30dcdcc5616df9b","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v2.1","peer_count":82,"percentile":18.52,"percentile_exact":{"denominator":27,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":645.68,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"nemotron-3-5-lightning:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":645.68},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"803b4920eccad03514634ebd3575a2d8","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"nemotron-3-5-lightning:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ec3633dbe4490f5d8334e3da32d471af","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":7.89,"percentile_exact":{"denominator":19,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.008410069634719542,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"nemotron-3-5-lightning:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008410069634719542},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"47de14e8ef5807902d31f4e62879bab3","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":78,"percentile":15.58,"percentile_exact":{"denominator":77,"numerator":1200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":506.85,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"nemotron-3-5-lightning:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":506.85},{"benchmark_family":"EnterpriseOps-Gym","benchmark_id":"b-a6fc65493a78d9050939","benchmark_label":"EnterpriseOps-Gym","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"80ae66f2b7393fe1f51bab7846dd0a10","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA run","peer_count":19,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; enterprise-ops-gym-aa","protocol_id":"b19a5badb616725bc5909b39","raw_value":0.18352730528200537,"reference_revision":"a23e9f106680ee4e45e01bc90992ce33b084abeb40b1d067a8d3f31f494ce21b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:enterprise-ops-gym-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:EnterpriseOps-Gym","source_record_id":"nemotron-3-5-lightning:enterprise-ops-gym-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18352730528200537},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c503c45f721e54502bd0c3c277c347cb","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"40a272cc2b1879bb5b0104ebb8833d18","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"current AA protocol","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.03,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"nemotron-3-5-lightning:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"nemotron-3-5-lightning:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.03},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8698ac189bdac0fe4a6655a8ed1ade73","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":11.32,"percentile_exact":{"denominator":53,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.14733333333333334,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.14733333333333334},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Nemotron 3.5 Lightning","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2184077575074c69e38c71abae562f14","included":true,"model_id":"m-9ff6170077842b96","original_version_label":"v1.1","peer_count":54,"percentile":52.83,"percentile_exact":{"denominator":53,"numerator":2800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6430805316653636,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6430805316653636}],"release_date":"2026-08-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":14.44,"score_exact":{"denominator":41274330401304,"numerator":596156995131155},"source_id":"nvidia/nemotron-3.5-lightning"},{"benchmark_count":13,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1095276,"numerator":17204225},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2000},"family":"GPQA Diamond","row_ids":["eebaf08b0e483d418dfe3796053cb2a9"],"score":33.9},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["b878256e74c2c14b269834f44b853cea"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":21,"numerator":100},"family":"HLE","row_ids":["19f9062134531b120e5a3beb1b827183"],"score":4.76},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":350},"family":"CritPt","row_ids":["1ce53974b2fd41332a137498132cd83b"],"score":6.86}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":33.9,"percentile_exact":{"denominator":59,"numerator":2000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","peer_count":64,"percentile":4.76,"percentile_exact":{"denominator":21,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01}],"score":15.71,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":7702695,"numerator":93810898},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":125},"family":"tau3-Banking","row_ids":["a7dc3e52deea9234fc190c75c4007564"],"score":9.62},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":500},"family":"GDPval-AA","row_ids":["72fabae9540ce0210203fd32c61249bb"],"score":6.17},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["c32f27bfb7b7fcf6e0fb4b8f49b0314a"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":175},"family":"AutomationBench","row_ids":["427f2b62a677b7b14b79a216cd5cce2a"],"score":9.21},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":300},"family":"AA-Briefcase","row_ids":["f3f72292d6ff516b21ceaec7da17a866"],"score":3.9}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":9.62,"percentile_exact":{"denominator":13,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","peer_count":82,"percentile":6.17,"percentile_exact":{"denominator":81,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":9.21,"percentile_exact":{"denominator":19,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":78,"percentile":3.9,"percentile_exact":{"denominator":77,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45}],"score":12.18,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":400},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":400},"family":"FrontierCode","row_ids":["e2d037eb56a1707481cab5a658eb199f"],"score":8.16}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"2.0","peer_count":50,"percentile":8.16,"percentile_exact":{"denominator":49,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101}],"score":8.16,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":3021,"numerator":62900},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":2,"numerator":75},"family":"IFBench","row_ids":["65c9161268988be090ca4411568e9e7b"],"score":37.5},{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":375},"family":"GDP.pdf","row_ids":["96fc6325ca8f16276f80447a5bb8a745"],"score":9.87},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":800},"family":"AA-Omniscience","row_ids":["2d5bfdd87d275a4c6177037eb5a3ff85","30b64e451b40a0bb848f166f0f44b6dd"],"score":15.09}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037}],"score":20.82,"weight":20}],"eligible":true,"id":"m-2c518178af26d1d4","in_window":true,"name":"North Mini Code","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-06-09","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":100,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"eebaf08b0e483d418dfe3796053cb2a9","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":33.9,"percentile_exact":{"denominator":59,"numerator":2000},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.756565656565657,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"north-mini-code:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.756565656565657},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"b878256e74c2c14b269834f44b853cea","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"19f9062134531b120e5a3beb1b827183","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"Diamond","peer_count":64,"percentile":4.76,"percentile_exact":{"denominator":21,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.042,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"north-mini-code:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.042},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1ce53974b2fd41332a137498132cd83b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2 Hard","peer_count":52,"percentile":6.86,"percentile_exact":{"denominator":51,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.01,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"claude-sonnet-5-5-medium:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7dc3e52deea9234fc190c75c4007564","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":9.62,"percentile_exact":{"denominator":13,"numerator":125},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0639175257731959,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"north-mini-code:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0639175257731959},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"72fabae9540ce0210203fd32c61249bb","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v2.1","peer_count":82,"percentile":6.17,"percentile_exact":{"denominator":81,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":318.62,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"north-mini-code:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":318.62},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c32f27bfb7b7fcf6e0fb4b8f49b0314a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"north-mini-code:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"427f2b62a677b7b14b79a216cd5cce2a","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":9.21,"percentile_exact":{"denominator":19,"numerator":175},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.014912858603711226,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"north-mini-code:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.014912858603711226},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f3f72292d6ff516b21ceaec7da17a866","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":78,"percentile":3.9,"percentile_exact":{"denominator":77,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":212.45,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"north-mini-code:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":212.45},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"e2d037eb56a1707481cab5a658eb199f","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"2.0","peer_count":50,"percentile":8.16,"percentile_exact":{"denominator":49,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0.0018760101010101,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0018760101010101},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"65c9161268988be090ca4411568e9e7b","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current 294-item protocol","peer_count":17,"percentile":37.5,"percentile_exact":{"denominator":2,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.575510204081633,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"north-mini-code:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.575510204081633},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"96fc6325ca8f16276f80447a5bb8a745","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"current AA protocol","peer_count":77,"percentile":9.87,"percentile_exact":{"denominator":38,"numerator":375},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.018,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"north-mini-code:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"north-mini-code:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"2d5bfdd87d275a4c6177037eb5a3ff85","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":16.98,"percentile_exact":{"denominator":53,"numerator":900},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.18366666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.18366666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"North Mini Code","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"30b64e451b40a0bb848f166f0f44b6dd","included":true,"model_id":"m-2c518178af26d1d4","original_version_label":"v1.1","peer_count":54,"percentile":13.21,"percentile_exact":{"denominator":53,"numerator":700},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.16496529195590037,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.16496529195590037}],"release_date":"2026-06-09","release_status":"published","release_verified":true,"result_updated_at":null,"score":14.19,"score_exact":{"denominator":3174948492408,"numerator":45043683141617},"source_id":"cohere/north-mini-code-1-0"},{"benchmark_count":14,"configuration":"公开协议汇总","coverage":5,"coverage_percent":100.0,"dimensions":[{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":67424575},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":2100},"family":"GPQA Diamond","row_ids":["13075042d3067db5be44e3af38eac3de"],"score":35.59},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["705b0d6767e14eda4ee6da31459437c6"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":850},"family":"HLE","row_ids":["48d08d1a4f13b1f74f0a43f0e70c78cb"],"score":13.49},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":800},"family":"CritPt","row_ids":["bfeb15a230a3fe93a4b8668ae5c71e33"],"score":15.69}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":35.59,"percentile_exact":{"denominator":59,"numerator":2100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","peer_count":52,"percentile":15.69,"percentile_exact":{"denominator":51,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025}],"score":20.52,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":2567565,"numerator":38748266},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":100},"family":"tau3-Banking","row_ids":["d4ac6d97f1d2caee1aac75fe86cb1d41"],"score":7.69},{"benchmark_version":"2.1","exact":{"denominator":27,"numerator":400},"family":"GDPval-AA","row_ids":["a7f8996adb7838234a9dd47a6d848a6c"],"score":14.81},{"benchmark_version":"4.0","exact":{"denominator":1,"numerator":32},"family":"Terminal-Bench","row_ids":["ed0927341bbf29454c5f15ec4d2315cd"],"score":32.0},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":250},"family":"AutomationBench","row_ids":["c9594ae7e63800079a6fe5e87536af10"],"score":13.16},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":600},"family":"AA-Briefcase","row_ids":["d623fd47b28047c5307a67919c5c6fae"],"score":7.79}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","peer_count":82,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":13.16,"percentile_exact":{"denominator":19,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":78,"percentile":7.79,"percentile_exact":{"denominator":77,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2}],"score":15.09,"weight":25},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["c8987663cd611e1b09bc6211e110c938"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":20.0,"evaluator_count":1,"exact":{"denominator":4028,"numerator":81325},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":25},"family":"IFBench","row_ids":["3f2a6e6ace968612fc72d2719d04f5ee"],"score":6.25},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":100},"family":"GDP.pdf","row_ids":["6612c73a65ff838439bc02c707904f4b"],"score":5.26},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":2600},"family":"AA-Omniscience","row_ids":["0ca7e9728f633cc599fad0ed12a6097b","862933c1b9c8be1d0cbb4bb130b222e9"],"score":49.06}],"family_count":3,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","peer_count":17,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625}],"score":20.19,"weight":20},{"effective_weight_percent":10.0,"evaluator_count":1,"exact":{"denominator":17,"numerator":50},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":17,"numerator":50},"family":"MMMU-Pro","row_ids":["b79c49748f47d291a424306d40dc6469"],"score":2.94}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532}],"score":2.94,"weight":10}],"eligible":true,"id":"m-8877a9675c79fcad","in_window":true,"name":"Command A+","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-20","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":101,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"13075042d3067db5be44e3af38eac3de","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":35.59,"percentile_exact":{"denominator":59,"numerator":2100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.760606060606061,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"command-a-plus:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.760606060606061},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"705b0d6767e14eda4ee6da31459437c6","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"command-a-plus:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"48d08d1a4f13b1f74f0a43f0e70c78cb","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"command-a-plus:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"bfeb15a230a3fe93a4b8668ae5c71e33","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2 Hard","peer_count":52,"percentile":15.69,"percentile_exact":{"denominator":51,"numerator":800},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0.025,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"command-a-plus:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.025},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d4ac6d97f1d2caee1aac75fe86cb1d41","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":7.69,"percentile_exact":{"denominator":13,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0597938144329897,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"command-a-plus:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0597938144329897},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a7f8996adb7838234a9dd47a6d848a6c","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v2.1","peer_count":82,"percentile":14.81,"percentile_exact":{"denominator":27,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":521.85,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"command-a-plus:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":521.85},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"ed0927341bbf29454c5f15ec4d2315cd","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"4.0","peer_count":76,"percentile":32.0,"percentile_exact":{"denominator":1,"numerator":32},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0.00505050505050505,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"command-a-plus:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.00505050505050505},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c9594ae7e63800079a6fe5e87536af10","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":13.16,"percentile_exact":{"denominator":19,"numerator":250},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.018104537047977383,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"command-a-plus:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.018104537047977383},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"d623fd47b28047c5307a67919c5c6fae","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":78,"percentile":7.79,"percentile_exact":{"denominator":77,"numerator":600},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":356.2,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"command-a-plus:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":356.2},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"c8987663cd611e1b09bc6211e110c938","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"command-a-plus:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"3f2a6e6ace968612fc72d2719d04f5ee","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current 294-item protocol","peer_count":17,"percentile":6.25,"percentile_exact":{"denominator":4,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.364625850340136,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"jt-35b-flash:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"jt-35b-flash:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.364625850340136},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"6612c73a65ff838439bc02c707904f4b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current AA protocol","peer_count":77,"percentile":5.26,"percentile_exact":{"denominator":19,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.01,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"command-a-plus:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.01},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"0ca7e9728f633cc599fad0ed12a6097b","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":1.89,"percentile_exact":{"denominator":53,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.09833333333333333,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.09833333333333333},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Command A+","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"862933c1b9c8be1d0cbb4bb130b222e9","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"v1.1","peer_count":54,"percentile":96.23,"percentile_exact":{"denominator":53,"numerator":5100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.8613678373382625,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"command-a-plus:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.8613678373382625},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Command A+","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"b79c49748f47d291a424306d40dc6469","included":true,"model_id":"m-8877a9675c79fcad","original_version_label":"current Standard 10","peer_count":35,"percentile":2.94,"percentile_exact":{"denominator":17,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.632369942196532,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"command-a-plus:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"command-a-plus:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.632369942196532}],"release_date":"2026-05-20","release_status":"published","release_verified":true,"result_updated_at":null,"score":13.85,"score_exact":{"denominator":15286789037520,"numerator":211678630100407},"source_id":"cohere/command-a-plus-05-2026"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3,"numerator":50},"families":[{"benchmark_version":"Diamond","exact":{"denominator":3,"numerator":50},"family":"GPQA Diamond","row_ids":["55c9779a030f53f08065b75a89097776"],"score":16.67}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2}],"score":16.67,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":3,"numerator":55},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":20},"family":"APEX-Agents","row_ids":["6f3782a0a7f13ec12e5eae491c79c282"],"score":20.0},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["96fb24279652563e7d2bc82b71d0fc52"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":3,"numerator":100},"family":"AA-Briefcase","row_ids":["5e2ad6a27f43522ccb21c4f280e4613e"],"score":33.33},{"benchmark_version":"1000 tasks","exact":{"denominator":1,"numerator":20},"family":"MCP Atlas","row_ids":["d57170302e825b662bfe4acca3fbd398"],"score":20.0}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6}],"score":18.33,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"DeepSWE","row_ids":["546ef51508bcfe588737e01c240c5015"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":17,"numerator":250},"families":[{"benchmark_version":"run 1.2.0","exact":{"denominator":17,"numerator":250},"family":"SimpleQA Verified","row_ids":["bcb4fb304c488f86e521408952f81841"],"score":14.71}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"score":14.71,"weight":20}],"eligible":true,"id":"m-5cf6d704eda6778e","in_window":true,"name":"GLM-5.2","official_evidence":{"artifact_sha256":"c6c90dc0946031acbba79d00549ddb918494022b454142617bed1e9492fdd97e","basis":"official_release_page","date":"2026-06-16","sha256":"c6c90dc0946031acbba79d00549ddb918494022b454142617bed1e9492fdd97e","url":"https://z.ai/blog/glm-5.2"},"primary_profile_id":"native","profile_id":"native","rank":102,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"55c9779a030f53f08065b75a89097776","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Diamond; fixed 198-question subset","peer_count":7,"percentile":16.67,"percentile_exact":{"denominator":3,"numerator":50},"profile_key":"native","protocol_description":"Official published comparison; GPQA Diamond; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"920ae7339db73f9e73534a08","raw_value":91.2,"reference_revision":"d23e33a9e09c477aefd3f07093d842ce0fe9e559f4cc249df07f61e00a4ddce1","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"91.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row2:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:GPQA Diamond:GPQA Diamond","source_record_id":"kimi-k3:table1:row2:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":91.2},{"benchmark_family":"APEX-Agents","benchmark_id":"b-047f2599e176070c08d9","benchmark_label":"APEX-Agents","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"6f3782a0a7f13ec12e5eae491c79c282","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current published protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; APEX-Agents; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"427a1a2a441195d7f27614d9","raw_value":35.6,"reference_revision":"dcf7e7e350460f3b91da013ef007e73f83a8c35b18324df4291b8df4e9b7e85c","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"35.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row28:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:APEX-Agents:APEX-Agents","source_record_id":"kimi-k3:table1:row28:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":35.6},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"96fb24279652563e7d2bc82b71d0fc52","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"published developer protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row AutomationBench; preserve page footnotes and task-specific settings","protocol_id":"afcaaccdf9ce24d5aa463683","raw_value":12.9,"reference_revision":"cd541571a4846f6dc785dfad562296e1d157308db37a2f04a2d84e7cac795afb","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"12.9","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row24:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AutomationBench:AutomationBench","source_record_id":"kimi-k3:table1:row24:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":12.9},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:27:19.715485+00:00","component_key":"score","configuration":"GLM-5.2 (Max)","dimension":"agents","direction":"higher_better","feed_id":"public-aa-briefcase-rubric","id":"5e2ad6a27f43522ccb21c4f280e4613e","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"v1.1","peer_count":19,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"aa-briefcase-rubric-reviewed-method-20261029","protocol_id":"66efeab01f142e89eb5e2362","raw_value":0.3606060606060606,"reference_revision":"fe9246b592e7047f4c9d7ee49c83140a24cafdf8f3f038a7b4c210a4cb535833","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"ba0e7c1f5b6e70c09d9a33f7ba0cf670630625c0edfde2d374b56fdeef8c0033","result_date":null,"source_cell":null,"source_id":"s-094d1db307438c4fd53b","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"e7382b3de2f8578cf1e3e313"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:briefcase-rubric:1.1","source_record_id":"e7382b3de2f8578cf1e3e313","source_url":"https://artificialanalysis.ai/evaluations/aa-briefcase","unit":"比例","value":0.3606060606060606},{"benchmark_family":"MCP Atlas","benchmark_id":"b-caa51c229f13b780d415","benchmark_label":"MCP-Atlas","benchmark_version":"1000 tasks","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"agents","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"d57170302e825b662bfe4acca3fbd398","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"Current 1000-task public suite; preserve evaluator protocol","peer_count":6,"percentile":20.0,"percentile_exact":{"denominator":1,"numerator":20},"profile_key":"native","protocol_description":"Official published comparison; MCP-Atlas; exact column GLM-5.2 (max). Footnotes and all competing runs retained in source.","protocol_id":"f6ed528afdd4dc3202de03d2","raw_value":82.6,"reference_revision":"fd35e5d83d6db24c27193cfedfeef16f459503999c47e197c45dade837d3fcb8","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"82.6","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row23:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MCP Atlas:MCP-Atlas","source_record_id":"kimi-k3:table1:row23:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":82.6},{"benchmark_family":"DeepSWE","benchmark_id":"b-56278ee9d68287e48fd9","benchmark_label":"DeepSWE","benchmark_version":"1.1","captured_at":"2026-11-06T03:28:33.115782+00:00","component_key":"score","configuration":"GLM-5.2 (max)","dimension":"coding","direction":"higher_better","feed_id":"public-release-e7129e9021b63e3a","id":"546ef51508bcfe588737e01c240c5015","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"1.1; explicit developer footnote","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Published page kimi-k3; row DeepSWE; preserve page footnotes and task-specific settings","protocol_id":"3b7e071f01bc65c78bc74f4a","raw_value":46.2,"reference_revision":"adfd809492c017867be98678b1526de57c4386269f69322879106342c3f2894e","research_artifact_sha256":"6135941d97bd8b92e412f6507d00ad7a9fa400c6a079bd8e73f3aa00d563d792","response_sha256":"61095c0a8d098a8359e45d87ac6abb84f43672ce262c0fc8dab489750621299e","result_date":null,"source_cell":"46.2","source_id":"s-b72308abea4f68de1b3f","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"kimi-k3:table1:row7:column6"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:DeepSWE:DeepSWE","source_record_id":"kimi-k3:table1:row7:column6","source_url":"https://huggingface.co/moonshotai/Kimi-K3","unit":"%","value":46.2},{"benchmark_family":"SimpleQA Verified","benchmark_id":"b-b6b12cc57cf11e4dea7e","benchmark_label":"SimpleQA Verified","benchmark_version":"run 1.2.0","captured_at":"2026-11-08T03:27:20.411678+00:00","component_key":"score","configuration":"glm-5.2_max","dimension":"language_information","direction":"higher_better","feed_id":"public-epoch-core-public-csv","id":"bcb4fb304c488f86e521408952f81841","included":true,"model_id":"m-5cf6d704eda6778e","original_version_label":"run 1.2.0","peer_count":35,"percentile":14.71,"percentile_exact":{"denominator":17,"numerator":250},"profile_key":"native","protocol_description":"epoch-simpleqa-reviewed-method-20261029-task-version-1.2.0","protocol_id":"4a5d6edc10aebf40572422c5","raw_value":0.342,"reference_revision":"7732d006cfbe666491ecac5c6368bcfb1f9f235ca9752514d0bedacbddff8958","research_artifact_sha256":"94e47149c0df662e59646f0f2d520ba95403081a944f643db794a1f20a50579b","response_sha256":"b6c09f0f3e6104a72fcbe6b97fac0b8d8da6967ef69001f0bc0bbde30208b8c8","result_date":"2026-08-27","result_timestamp":"2026-08-27T19:36:05.000Z","source_cell":null,"source_id":"s-e78acdd5f50b212b8328","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"1942340bdaaf42991a1babd9"},"source_name":"Epoch AI","source_protocol_key":"epoch-simpleqa-reviewed-method-task-version-1.2.0","source_record_id":"1942340bdaaf42991a1babd9","source_url":"https://epoch.ai/data/benchmarks.csv","unit":"比例","value":0.342}],"release_date":"2026-06-16","release_status":"published","release_verified":true,"result_updated_at":"2026-08-27","score":12.99,"score_exact":{"denominator":102,"numerator":1325},"source_id":"zhipuai/glm-5.2"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":2,"exact":{"denominator":2394,"numerator":20875},"families":[{"benchmark_version":"0.1.0","exact":{"denominator":19,"numerator":75},"family":"Terminal-Bench-Science","row_ids":["80525908530cba67da8858ed258e1742"],"score":3.95},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":850},"family":"HLE","row_ids":["c6b63609fd0ed9a9d1cc3300620f0885"],"score":13.49}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051}],"score":8.72,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":1881,"numerator":23195},"families":[{"benchmark_version":"2.1","exact":{"denominator":9,"numerator":100},"family":"GDPval-AA","row_ids":["21dd109a3c14f1e9b389fa0a30a5ee0f"],"score":11.11},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["179a39191c3950939aafcb199439b020"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":300},"family":"AutomationBench","row_ids":["43181abb2675b0ef85c7558935e2c229"],"score":15.79},{"benchmark_version":"1.1","exact":{"denominator":11,"numerator":100},"family":"AA-Briefcase","row_ids":["c4bef304832e9ad8f150cf2ae6ad2c77"],"score":9.09}],"family_count":4,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","peer_count":82,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":15.79,"percentile_exact":{"denominator":19,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","peer_count":78,"percentile":9.09,"percentile_exact":{"denominator":11,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49}],"score":12.33,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":350},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":350},"family":"GDP.pdf","row_ids":["8f8d77f644fefbdcb7d35712e527c726"],"score":18.42}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036}],"score":18.42,"weight":20}],"eligible":true,"id":"m-b9d9f5989f59e750","in_window":true,"name":"Mercury 2.5","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-08","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":103,"records":[{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:28:53.336510+00:00","component_key":"score","configuration":"Mercury 2.5 / Mini-SWE-agent","dimension":"reasoning","direction":"higher_better","feed_id":"public-vals-terminalbenchscience","id":"80525908530cba67da8858ed258e1742","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"0.1.0","peer_count":39,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","protocol_id":"dbbbcaf695179b55af0fc57c","raw_value":0.0,"reference_revision":"63132d29fa4ebf3b7fbbf94239ba8e3db4666335ed2553c88a03e2feb59d51c6","research_artifact_sha256":"6d82d2fadaab57f797383e94997e64ace6d063ba8c34b6d1033d5d7e3568ebb8","response_sha256":"e7b205eb01a0ff83fd02009c8a0f5ef6138d1409df94817b1fbf3739f4dc2d93","result_date":null,"source_cell":"0.00%","source_id":"s-eb560d0cf6499ef1c9f9","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"twelve-vals-terminal-science:row40"},"source_name":"Vals AI","source_protocol_key":"Vals:Current displayed evaluator row. Mini-SWE-agent. Scores are not pooled across harnesses; source methodology and repeat counts retained.","source_record_id":"twelve-vals-terminal-science:row40","source_url":"https://www.vals.ai/benchmarks/terminal-bench-science","unit":"%","value":0.0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c6b63609fd0ed9a9d1cc3300620f0885","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"Diamond","peer_count":64,"percentile":13.49,"percentile_exact":{"denominator":63,"numerator":850},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.051,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"mercury-2-5:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.051},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"21dd109a3c14f1e9b389fa0a30a5ee0f","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v2.1","peer_count":82,"percentile":11.11,"percentile_exact":{"denominator":9,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":481.79,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"mercury-2-5:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":481.79},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"179a39191c3950939aafcb199439b020","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"mercury-2-5:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"43181abb2675b0ef85c7558935e2c229","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":15.79,"percentile_exact":{"denominator":19,"numerator":300},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.023541080303287186,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"mercury-2-5:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.023541080303287186},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"c4bef304832e9ad8f150cf2ae6ad2c77","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"v1.1","peer_count":78,"percentile":9.09,"percentile_exact":{"denominator":11,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":378.49,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"mercury-2-5:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":378.49},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Mercury 2.5","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8f8d77f644fefbdcb7d35712e527c726","included":true,"model_id":"m-b9d9f5989f59e750","original_version_label":"current AA protocol","peer_count":77,"percentile":18.42,"percentile_exact":{"denominator":19,"numerator":350},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.036,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"mercury-2-5:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"mercury-2-5:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.036}],"release_date":"2026-09-08","release_status":"published","release_verified":true,"result_updated_at":null,"score":12.78,"score_exact":{"denominator":40964,"numerator":523575},"source_id":"inception/mercury-2.5"},{"benchmark_count":9,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":2,"exact":{"denominator":531,"numerator":13000},"families":[{"benchmark_version":"2026","exact":{"denominator":3,"numerator":100},"family":"AIME","row_ids":["072b96c67ae19a20ee48a971117995c0"],"score":33.33},{"benchmark_version":"500-item subset","exact":{"denominator":3,"numerator":100},"family":"MATH-500","row_ids":["158a7a0bcbb4fdee990dff1e880fb156"],"score":33.33},{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":400},"family":"GPQA Diamond","row_ids":["249b24c9d6325f053552fb911f7f4a84"],"score":6.78}],"family_count":3,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","included":true,"model_id":"m-161fc8601724c921","original_version_label":"2026","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":6.78,"percentile_exact":{"denominator":59,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313}],"score":24.48,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"4","exact":{"denominator":1,"numerator":0},"family":"BFCL","row_ids":["602ba59a6dfe14df55137c17751e7666"],"score":0.0}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","included":true,"model_id":"m-161fc8601724c921","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5}],"score":0.0,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"release_v6","exact":{"denominator":1,"numerator":0},"family":"LiveCodeBench","row_ids":["445a31daff84293600dd4a2900cbf97d"],"score":0.0}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8}],"score":0.0,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":2,"exact":{"denominator":48,"numerator":775},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":4,"numerator":125},"family":"IFBench","row_ids":["950f9e2f9181b2c99ee5f8151838596d"],"score":31.25},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"Multi-IF","row_ids":["5b52e92a8499178ddfa84c801132f2f4"],"score":0.0},{"benchmark_version":"current maintainer dataset","exact":{"denominator":3,"numerator":100},"family":"IFEval","row_ids":["0c588d45b55a074bc2915de39e2cf5ff"],"score":33.33},{"benchmark_version":"current maintainer dataset","exact":{"denominator":1,"numerator":0},"family":"MMLU-Pro","row_ids":["0129fdb380d08e0198547b85719a77cd"],"score":0.0}],"family_count":4,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1}],"score":16.15,"weight":20}],"eligible":true,"id":"m-161fc8601724c921","in_window":true,"name":"LFM2.5-8B-A1B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-28","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":104,"records":[{"benchmark_family":"AIME","benchmark_id":"b-2f883043a8949277d5d0","benchmark_label":"AIME 2026","benchmark_version":"2026","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"072b96c67ae19a20ee48a971117995c0","included":true,"model_id":"m-161fc8601724c921","original_version_label":"2026","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; AIME 2026; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"97974a29c366bdc315df37df","raw_value":56.7,"reference_revision":"ac3f70834adc73f25f08d30ec0ffb8f226b56c4ef60b3caee25f1b9ec24350f2","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"56.7","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row11:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:AIME:AIME 2026","source_record_id":"twelve-minicpm2-card:table0:row11:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":56.7},{"benchmark_family":"MATH-500","benchmark_id":"b-613da685e272977e9e6e","benchmark_label":"MATH-500","benchmark_version":"500-item subset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"158a7a0bcbb4fdee990dff1e880fb156","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current fixed 500-problem subset; preserve evaluator protocol","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; MATH-500; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"a181f644118455cd59af445a","raw_value":93.2,"reference_revision":"fd9ccbc4ee7d1196610c82f80af621f3056ab3a7002634c3da1045d0ff838489","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"93.2","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row13:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MATH-500:MATH-500","source_record_id":"twelve-minicpm2-card:table0:row13:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":93.2},{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"249b24c9d6325f053552fb911f7f4a84","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":6.78,"percentile_exact":{"denominator":59,"numerator":400},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.513131313131313,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"lfm2-5-8b-a1b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.513131313131313},{"benchmark_family":"BFCL","benchmark_id":"b-bbbb771a6f0277cf8f66","benchmark_label":"BFCLv4","benchmark_version":"4","captured_at":"2026-11-06T03:27:23.162587+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"agents","direction":"higher_better","feed_id":"public-release-21fb8d989d891cfa","id":"602ba59a6dfe14df55137c17751e7666","included":true,"model_id":"m-161fc8601724c921","original_version_label":"v4","peer_count":2,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; BFCLv4; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"5c8ad28d0b22630243378008","raw_value":48.5,"reference_revision":"2e4a1ea426c62b1353b143c11edb08bb93fa5ad274dc1dc47e015edd763c5b02","research_artifact_sha256":"7a4e532090b51cfae712420a19dae86b9a9eebb5719e173381491948c7d24267","response_sha256":"0b7bc24451aae276680894b241e9467a66badc7eea38be9a4993eee8d3ab7716","result_date":null,"source_cell":"48.50","source_id":"s-6bc244425c066f47875a","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-liquid8-card:table4:row10:column2"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:BFCL:BFCLv4","source_record_id":"twelve-liquid8-card:table4:row10:column2","source_url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B","unit":"%","value":48.5},{"benchmark_family":"LiveCodeBench","benchmark_id":"b-c80b91404411028a7b24","benchmark_label":"LiveCodeBench v6","benchmark_version":"release_v6","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"coding","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"445a31daff84293600dd4a2900cbf97d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"release_v6; window and repeat count kept in protocol","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; LiveCodeBench v6; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9b21b259ed1433cdf48c1913","raw_value":39.8,"reference_revision":"5636a86c1bd98cc40850f4f5649c28b3d68c75dee25c032052c94a5c02bae329","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"39.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row4:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:LiveCodeBench:LiveCodeBench v6","source_record_id":"twelve-minicpm2-card:table0:row4:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":39.8},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"950f9e2f9181b2c99ee5f8151838596d","included":true,"model_id":"m-161fc8601724c921","original_version_label":"current 294-item protocol","peer_count":17,"percentile":31.25,"percentile_exact":{"denominator":4,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.556462585034014,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"lfm2-5-8b-a1b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"lfm2-5-8b-a1b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.556462585034014},{"benchmark_family":"Multi-IF","benchmark_id":"b-42d7cc897e4d47cf8b12","benchmark_label":"Multi-IF","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"5b52e92a8499178ddfa84c801132f2f4","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Meta dataset","peer_count":4,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; Multi-IF; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"9400ddb5b35b4ffd1dad74b3","raw_value":71.4,"reference_revision":"b666715c74c9e09e19a163a1fa3f67617cc91d6408c7ff1ee40c14b39d5223a9","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"71.4","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row17:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:Multi-IF:Multi-IF","source_record_id":"twelve-minicpm2-card:table0:row17:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":71.4},{"benchmark_family":"IFEval","benchmark_id":"b-4ff0d64c85eab1c76764","benchmark_label":"IFEval","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0c588d45b55a074bc2915de39e2cf5ff","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current original Google dataset","peer_count":4,"percentile":33.33,"percentile_exact":{"denominator":3,"numerator":100},"profile_key":"native","protocol_description":"Official published comparison; IFEval; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"b7479430c87ac4524f0a5b12","raw_value":90.8,"reference_revision":"fa9cb760c18105fb2029d0c3ca93594441605ec0993efcb61d2baab293f24e8b","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"90.8","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row16:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:IFEval:IFEval","source_record_id":"twelve-minicpm2-card:table0:row16:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":90.8},{"benchmark_family":"MMLU-Pro","benchmark_id":"b-a774e6af7c948cf88133","benchmark_label":"MMLU-Pro","benchmark_version":"current maintainer dataset","captured_at":"2026-11-06T03:28:18.078893+00:00","component_key":"score","configuration":"LFM2.5-8B-A1B","dimension":"language_information","direction":"higher_better","feed_id":"public-release-6db7a09f6b2d61e8","id":"0129fdb380d08e0198547b85719a77cd","included":true,"model_id":"m-161fc8601724c921","original_version_label":"Current maintainer dataset; evaluator-specific protocol","peer_count":5,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Official published comparison; MMLU-Pro; exact column LFM2.5-8B-A1B. Footnotes and all competing runs retained in source.","protocol_id":"dff05c56b6d8c6ac1cddbc91","raw_value":63.1,"reference_revision":"75d8ec3bbf547c7d8a5f330a313b0937ebec654730bcb0590957b3f09bbeee80","research_artifact_sha256":"7cc48bdc8a49c6dccfea1ec36280098901d9fb41a236453e07a2457d48763403","response_sha256":"439c2f201b2a89095dcf79a5613b266aba2dd2acb7f7bcc17a6bfd668097447b","result_date":null,"source_cell":"63.1","source_id":"s-2d4bd7b6b48d97ab2124","source_kind":"official_release_comparison","source_locator":{"image_url":null,"record":"twelve-minicpm2-card:table0:row19:column9"},"source_name":"huggingface.co","source_protocol_key":"official:huggingface.co:MMLU-Pro:MMLU-Pro","source_record_id":"twelve-minicpm2-card:table0:row19:column9","source_url":"https://huggingface.co/openbmb/MiniCPM5-2B","unit":"%","value":63.1}],"release_date":"2026-05-28","release_status":"published","release_verified":true,"result_updated_at":null,"score":10.39,"score_exact":{"denominator":38232,"numerator":397175},"source_id":"published/aa:lfm2-5-8b-a1b"},{"benchmark_count":12,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":3285828,"numerator":28072925},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":750},"family":"GPQA Diamond","row_ids":["80f85d325d6536dbd1603ecf4678090b"],"score":12.71},{"benchmark_version":"0.1.0","exact":{"denominator":13,"numerator":225},"family":"Terminal-Bench-Science","row_ids":["10a24da4314ddad1e5487011d9ce22be"],"score":17.31},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":200},"family":"HLE","row_ids":["1c035cc3ccd2789e7031fcc995e521f7"],"score":3.17},{"benchmark_version":"2 Hard","exact":{"denominator":51,"numerator":50},"family":"CritPt","row_ids":["8e1dd162a62fd85fef38ef1d97e4e9bc"],"score":0.98}],"family_count":4,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","peer_count":64,"percentile":3.17,"percentile_exact":{"denominator":63,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":8.54,"weight":25},{"effective_weight_percent":27.78,"evaluator_count":1,"exact":{"denominator":1540539,"numerator":18519664},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":175},"family":"tau3-Banking","row_ids":["a1e714634bbaa4b289e4726c4bb2c9e9"],"score":13.46},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":1000},"family":"GDPval-AA","row_ids":["5be65e1004d9e6f482549aefe65a50a4"],"score":12.35},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["f435dfe97ff078f0f04715c2a33b7561"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":275},"family":"AutomationBench","row_ids":["e3a197e716166eb9edc4c46f49e52295"],"score":14.47},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":500},"family":"AA-Briefcase","row_ids":["2e097f3022508687e5c06477d01a5253"],"score":6.49}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":13.46,"percentile_exact":{"denominator":13,"numerator":175},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","peer_count":82,"percentile":12.35,"percentile_exact":{"denominator":81,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":78,"percentile":6.49,"percentile_exact":{"denominator":77,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41}],"score":12.02,"weight":25},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":49,"numerator":150},"families":[{"benchmark_version":"2.0","exact":{"denominator":49,"numerator":150},"family":"FrontierCode","row_ids":["6a0e7ddb2b9af667874b506e526b6d54"],"score":3.06}],"family_count":1,"key":"coding","label":"编程","records":[{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0}],"score":3.06,"weight":20},{"effective_weight_percent":22.22,"evaluator_count":1,"exact":{"denominator":2014,"numerator":35525},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":25},"family":"GDP.pdf","row_ids":["8eaabce7efa57428ec82adf89ba50dbc"],"score":1.32},{"benchmark_version":"1.1","exact":{"denominator":53,"numerator":1800},"family":"AA-Omniscience","row_ids":["4a7b9f5169709891b5917ee76e2a8e1a","5f020104c70b0364755cf0f561bff098"],"score":33.96}],"family_count":2,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647}],"score":17.64,"weight":20}],"eligible":true,"id":"m-1d98b8595cf1927d","in_window":true,"name":"Granite 4.2 8B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-08-25","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":105,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"80f85d325d6536dbd1603ecf4678090b","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"granite-4-2-8b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"Terminal-Bench-Science","benchmark_id":"b-c2390eb36c7cc7563feb","benchmark_label":"Terminal-Bench-Science","benchmark_version":"0.1.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"10a24da4314ddad1e5487011d9ce22be","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"0.1","peer_count":53,"percentile":17.31,"percentile_exact":{"denominator":13,"numerator":225},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminal-bench-science","protocol_id":"a09e264ffddc775882a3b90f","raw_value":0,"reference_revision":"0e0e6935e49877effcb6f6aec9249f46ba95308a59834356cee71f08e200cea0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:terminal-bench-science"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench-Science","source_record_id":"claude-sonnet-5-5-medium:terminal-bench-science","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"1c035cc3ccd2789e7031fcc995e521f7","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"Diamond","peer_count":64,"percentile":3.17,"percentile_exact":{"denominator":63,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.04,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"granite-4-2-8b:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.04},{"benchmark_family":"CritPt","benchmark_id":"b-df5cac92faa6d4335f3c","benchmark_label":"CritPt","benchmark_version":"2 Hard","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"8e1dd162a62fd85fef38ef1d97e4e9bc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2 Hard","peer_count":52,"percentile":0.98,"percentile_exact":{"denominator":51,"numerator":50},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; critpt-v2-hard","protocol_id":"ad076bfd835af583f4130282","raw_value":0,"reference_revision":"1befbad866117d8779573ae34aca3eb2ae192ff2f9f28076654410999c6a98b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:critpt-v2-hard"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:CritPt","source_record_id":"granite-4-2-8b:critpt-v2-hard","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"a1e714634bbaa4b289e4726c4bb2c9e9","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":13.46,"percentile_exact":{"denominator":13,"numerator":175},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0762886597938144,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"granite-4-2-8b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0762886597938144},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"5be65e1004d9e6f482549aefe65a50a4","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v2.1","peer_count":82,"percentile":12.35,"percentile_exact":{"denominator":81,"numerator":1000},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":498.74,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"granite-4-2-8b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":498.74},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"f435dfe97ff078f0f04715c2a33b7561","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"granite-4-2-8b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e3a197e716166eb9edc4c46f49e52295","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":14.47,"percentile_exact":{"denominator":19,"numerator":275},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.020571262692004582,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"granite-4-2-8b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.020571262692004582},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"2e097f3022508687e5c06477d01a5253","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":78,"percentile":6.49,"percentile_exact":{"denominator":77,"numerator":500},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":310.41,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"granite-4-2-8b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":310.41},{"benchmark_family":"FrontierCode","benchmark_id":"b-0d3dc041c8afc894da74","benchmark_label":"FrontierCode","benchmark_version":"2.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"coding","direction":"higher_better","feed_id":"public-aa-current","id":"6a0e7ddb2b9af667874b506e526b6d54","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"2.0","peer_count":50,"percentile":3.06,"percentile_exact":{"denominator":49,"numerator":150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; frontiercode-2-0","protocol_id":"7455c4b3b3db4769bfc4048f","raw_value":0,"reference_revision":"ed7a050f55b73237374c6261c4a773246388e21e832217a743c52de554fcb76e","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:frontiercode-2-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:FrontierCode","source_record_id":"claude-sonnet-5-5-medium:frontiercode-2-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"8eaabce7efa57428ec82adf89ba50dbc","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"current AA protocol","peer_count":77,"percentile":1.32,"percentile_exact":{"denominator":19,"numerator":25},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.006,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"granite-4-2-8b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"granite-4-2-8b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.006},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"accuracy","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"4a7b9f5169709891b5917ee76e2a8e1a","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":7.55,"percentile_exact":{"denominator":53,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-accuracy","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.11766666666666667,"reference_revision":"e0daf63e0f5039c8d65edcc76832456d73f5c9dae048ad63dd3cf707846ba9d2","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-accuracy"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-accuracy","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.11766666666666667},{"benchmark_family":"AA-Omniscience","benchmark_id":"b-fabd054a531c3d3b4156","benchmark_label":"AA-Omniscience","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"non_hallucination","configuration":"Granite 4.2 8B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"5f020104c70b0364755cf0f561bff098","included":true,"model_id":"m-1d98b8595cf1927d","original_version_label":"v1.1","peer_count":54,"percentile":60.38,"percentile_exact":{"denominator":53,"numerator":3200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; omniscience-non-hallucination","protocol_id":"57ce78dfdf41b4d0acd6e37d","raw_value":0.6777483944087647,"reference_revision":"2606d1d63180325444386c24b08fd7ac18eef6707e954e957fde4a9d24154fe8","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"claude-sonnet-5-5-medium:omniscience-non-hallucination"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Omniscience","source_record_id":"claude-sonnet-5-5-medium:omniscience-non-hallucination","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.6777483944087647}],"release_date":"2026-08-25","release_status":"published","release_verified":true,"result_updated_at":null,"score":10.31,"score_exact":{"denominator":41274330401304,"numerator":425645973328915},"source_id":"ibm/granite-4.2-8b"},{"benchmark_count":8,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":3717,"numerator":57550},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":750},"family":"GPQA Diamond","row_ids":["63aced451500a12a72f3e068bcd6ff1f"],"score":12.71},{"benchmark_version":"Diamond","exact":{"denominator":63,"numerator":1150},"family":"HLE","row_ids":["04be87f5b8e53109685297a0e0936bc5"],"score":18.25}],"family_count":2,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057}],"score":15.48,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":118503,"numerator":588163},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":1,"numerator":0},"family":"tau3-Banking","row_ids":["09489dd17969a18613dce0aff6c534ba"],"score":0.0},{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":400},"family":"GDPval-AA","row_ids":["0753db930b01dc3794424662db360ca1"],"score":4.94},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["732ce0d4c4b93e00640afbdf97e81cd3"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":75},"family":"AutomationBench","row_ids":["b4943be7200f5ac57de9f15aacbca48d"],"score":3.95},{"benchmark_version":"1.1","exact":{"denominator":77,"numerator":200},"family":"AA-Briefcase","row_ids":["dda521b5d7d0c5ce2d8db12c4ea78f4c"],"score":2.6}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","peer_count":82,"percentile":4.94,"percentile_exact":{"denominator":81,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","peer_count":78,"percentile":2.6,"percentile_exact":{"denominator":77,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94}],"score":4.96,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":19,"numerator":125},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":19,"numerator":125},"family":"GDP.pdf","row_ids":["d70ece05a7cd9206504e0831b954b698"],"score":6.58}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012}],"score":6.58,"weight":20}],"eligible":true,"id":"m-7990b74b056a0b9a","in_window":true,"name":"Celeris-1","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-07-24","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":106,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"63aced451500a12a72f3e068bcd6ff1f","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":12.71,"percentile_exact":{"denominator":59,"numerator":750},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.631313131313131,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"celeris-1:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.631313131313131},{"benchmark_family":"HLE","benchmark_id":"b-d1a65cedc5803489d8eb","benchmark_label":"HLE","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"04be87f5b8e53109685297a0e0936bc5","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"Diamond","peer_count":64,"percentile":18.25,"percentile_exact":{"denominator":63,"numerator":1150},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; hle-diamond","protocol_id":"09ebba97991aa218642551d3","raw_value":0.057,"reference_revision":"87078bc29d8685596915f3a30652f498bc762de5dfb8dc09b2bab58375b2d6a1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:hle-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:HLE","source_record_id":"celeris-1:hle-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.057},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"09489dd17969a18613dce0aff6c534ba","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0391752577319588,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"celeris-1:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0391752577319588},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"0753db930b01dc3794424662db360ca1","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v2.1","peer_count":82,"percentile":4.94,"percentile_exact":{"denominator":81,"numerator":400},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":309.46,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"celeris-1:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":309.46},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"732ce0d4c4b93e00640afbdf97e81cd3","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"celeris-1:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"b4943be7200f5ac57de9f15aacbca48d","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":3.95,"percentile_exact":{"denominator":19,"numerator":75},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.004158035881179356,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"celeris-1:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.004158035881179356},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"dda521b5d7d0c5ce2d8db12c4ea78f4c","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"v1.1","peer_count":78,"percentile":2.6,"percentile_exact":{"denominator":77,"numerator":200},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":151.94,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"celeris-1:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":151.94},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"Celeris-1","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"d70ece05a7cd9206504e0831b954b698","included":true,"model_id":"m-7990b74b056a0b9a","original_version_label":"current AA protocol","peer_count":77,"percentile":6.58,"percentile_exact":{"denominator":19,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.012,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"celeris-1:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"celeris-1:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.012}],"release_date":"2026-07-24","release_status":"published","release_verified":true,"result_updated_at":null,"score":9.18,"score_exact":{"denominator":13983354,"numerator":128393905},"source_id":"published/aa:celeris-1"},{"benchmark_count":7,"configuration":"公开协议汇总","coverage":3,"coverage_percent":60.0,"dimensions":[{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":59,"numerator":100},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":100},"family":"GPQA Diamond","row_ids":["990a7e694606915f66b15d0a96c06abb"],"score":1.69}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","included":true,"model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":1.69,"percentile_exact":{"denominator":59,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293}],"score":1.69,"weight":25},{"effective_weight_percent":35.71,"evaluator_count":1,"exact":{"denominator":39,"numerator":119},"families":[{"benchmark_version":"1.0.1","exact":{"denominator":13,"numerator":25},"family":"tau3-Banking","row_ids":["57be0e2f641053db7912c66d73e8ba7b"],"score":1.92},{"benchmark_version":"2.1","exact":{"denominator":1,"numerator":0},"family":"GDPval-AA","row_ids":["710c23bbfe14fa2ea7bd1f3a24100a69"],"score":0.0},{"benchmark_version":"4.0","exact":{"denominator":3,"numerator":40},"family":"Terminal-Bench","row_ids":["7ed61dbea4d39e1cec33826bfcfe4762"],"score":13.33},{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"AutomationBench","row_ids":["e01a487a4aa0d6109e212abc8d8c200e"],"score":0.0},{"benchmark_version":"1.1","exact":{"denominator":1,"numerator":0},"family":"AA-Briefcase","row_ids":["3f053e6933f1e5ef77b0bf87fdfb5476"],"score":0.0}],"family_count":5,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","included":true,"model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":1.92,"percentile_exact":{"denominator":13,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v2.1","peer_count":82,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","included":true,"model_id":"m-928c410ba1082113","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","included":true,"model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v1.1","peer_count":78,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0}],"score":3.05,"weight":25},{"effective_weight_percent":28.57,"evaluator_count":1,"exact":{"denominator":38,"numerator":125},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":38,"numerator":125},"family":"GDP.pdf","row_ids":["b9fa019c4e1362fc93bc4088a9dc8dbc"],"score":3.29}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","included":true,"model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008}],"score":3.29,"weight":20}],"eligible":true,"id":"m-928c410ba1082113","in_window":true,"name":"K2 Horizon 0.9B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-09-03","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":107,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"990a7e694606915f66b15d0a96c06abb","included":true,"model_id":"m-928c410ba1082113","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":1.69,"percentile_exact":{"denominator":59,"numerator":100},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.292929292929293,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"k2-horizon-0-9b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.292929292929293},{"benchmark_family":"tau3-Banking","benchmark_id":"b-2328e41cefc33a2cab70","benchmark_label":"tau3-Banking","benchmark_version":"1.0.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"57be0e2f641053db7912c66d73e8ba7b","included":true,"model_id":"m-928c410ba1082113","original_version_label":"upstream v1.0.1 / 97 tasks / current AA pass@1","peer_count":53,"percentile":1.92,"percentile_exact":{"denominator":13,"numerator":25},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section tau3-banking","protocol_id":"416e531f10637380f0b1b24d","raw_value":0.0494845360824742,"reference_revision":"5293ba854bda2bc67b56b45771a75739d59cd11e74e2f396f9b34376dbad39e9","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:tau3-banking"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:tau3-Banking","source_record_id":"k2-horizon-0-9b:tau3-banking","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.0494845360824742},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"710c23bbfe14fa2ea7bd1f3a24100a69","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v2.1","peer_count":82,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-402.18,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"k2-horizon-0-9b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-402.18},{"benchmark_family":"Terminal-Bench","benchmark_id":"b-52b1cf5ccbedf0cff523","benchmark_label":"Terminal-Bench","benchmark_version":"4.0","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"7ed61dbea4d39e1cec33826bfcfe4762","included":true,"model_id":"m-928c410ba1082113","original_version_label":"4.0","peer_count":76,"percentile":13.33,"percentile_exact":{"denominator":3,"numerator":40},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; terminalbench-4-0","protocol_id":"daf4f954f79ad943ce53905c","raw_value":0,"reference_revision":"6a56af6dbe0ee4f0e6d8219f4788c253c320558b3cd93502c02b922226abd919","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:terminalbench-4-0"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:Terminal-Bench","source_record_id":"k2-horizon-0-9b:terminalbench-4-0","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0},{"benchmark_family":"AutomationBench","benchmark_id":"b-55877e849031d5ecca0c","benchmark_label":"AutomationBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"e01a487a4aa0d6109e212abc8d8c200e","included":true,"model_id":"m-928c410ba1082113","original_version_label":"AA 1.0.6 held-out","peer_count":77,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; automationbench-aa","protocol_id":"5434f491d6ad286fe20f125d","raw_value":0.002038009788550524,"reference_revision":"0df8245d8a3df62f6334a99faa9363d8951c1a3dfd3688cd0aa10e12da1bd049","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:automationbench-aa"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AutomationBench","source_record_id":"k2-horizon-0-9b:automationbench-aa","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.002038009788550524},{"benchmark_family":"AA-Briefcase","benchmark_id":"b-5b1bd3134f557e876a1d","benchmark_label":"AA-Briefcase","benchmark_version":"1.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"3f053e6933f1e5ef77b0bf87fdfb5476","included":true,"model_id":"m-928c410ba1082113","original_version_label":"v1.1","peer_count":78,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; aa-briefcase","protocol_id":"e0d5f43b6db1c39db024366a","raw_value":0,"reference_revision":"c4b3e738ffc072d058db25ce65309476d788cb333d273515fe4f1391a039b9e3","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:aa-briefcase"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:AA-Briefcase","source_record_id":"k2-horizon-0-9b:aa-briefcase","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":0},{"benchmark_family":"GDP.pdf","benchmark_id":"b-4d63308d53e1487db1c8","benchmark_label":"GDP.pdf","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"K2 Horizon 0.9B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"b9fa019c4e1362fc93bc4088a9dc8dbc","included":true,"model_id":"m-928c410ba1082113","original_version_label":"current AA protocol","peer_count":77,"percentile":3.29,"percentile_exact":{"denominator":38,"numerator":125},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdp-pdf","protocol_id":"82ccf77be7c53102e3bec8a8","raw_value":0.008,"reference_revision":"3a0e1f8043eabf05fde80ed3fb8e0e6694c515f0301320145e468b0038554da1","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"k2-horizon-0-9b:gdp-pdf"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDP.pdf","source_record_id":"k2-horizon-0-9b:gdp-pdf","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.008}],"release_date":"2026-09-03","release_status":"published","release_verified":true,"result_updated_at":null,"score":2.63,"score_exact":{"denominator":612066,"numerator":1612745},"source_id":"published/aa:k2-horizon-0-9b"},{"benchmark_count":4,"configuration":"公开协议汇总","coverage":4,"coverage_percent":80.0,"dimensions":[{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":59,"numerator":200},"families":[{"benchmark_version":"Diamond","exact":{"denominator":59,"numerator":200},"family":"GPQA Diamond","row_ids":["c0db3282b4d173e68108771ffab82d19"],"score":3.39}],"family_count":1,"key":"reasoning","label":"推理求解","records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":3.39,"percentile_exact":{"denominator":59,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505}],"score":3.39,"weight":25},{"effective_weight_percent":31.25,"evaluator_count":1,"exact":{"denominator":81,"numerator":100},"families":[{"benchmark_version":"2.1","exact":{"denominator":81,"numerator":100},"family":"GDPval-AA","row_ids":["32822274d56779c973cdcf3ccb12841f"],"score":1.23}],"family_count":1,"key":"agents","label":"工具与任务执行","records":[{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","peer_count":82,"percentile":1.23,"percentile_exact":{"denominator":81,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32}],"score":1.23,"weight":25},{"effective_weight_percent":25.0,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"current published task suite","exact":{"denominator":1,"numerator":0},"family":"IFBench","row_ids":["e359de5ff096ac45c506a2baeb714ed7"],"score":0.0}],"family_count":1,"key":"language_information","label":"语言与信息处理","records":[{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","peer_count":17,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551}],"score":0.0,"weight":20},{"effective_weight_percent":12.5,"evaluator_count":1,"exact":{"denominator":1,"numerator":0},"families":[{"benchmark_version":"Standard 10","exact":{"denominator":1,"numerator":0},"family":"MMMU-Pro","row_ids":["7a567c9ea42d069b6cee3f34d322f5a9"],"score":0.0}],"family_count":1,"key":"multimodal_understanding","label":"多模态理解","records":[{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087}],"score":0.0,"weight":10}],"eligible":true,"id":"m-0cd6b9e2d04f2c30","in_window":true,"name":"MiniCPM-V 4.6 1.3B","official_evidence":{"artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","basis":"authoritative_explicit_release_record","date":"2026-05-11","sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","url":"https://artificialanalysis.ai/leaderboards/models"},"primary_profile_id":"native","profile_id":"native","rank":108,"records":[{"benchmark_family":"GPQA Diamond","benchmark_id":"b-644121bda56f6eaf2dc0","benchmark_label":"GPQA Diamond","benchmark_version":"Diamond","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"reasoning","direction":"higher_better","feed_id":"public-aa-current","id":"c0db3282b4d173e68108771ffab82d19","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"Diamond / 198 items / current AA pass@1","peer_count":60,"percentile":3.39,"percentile_exact":{"denominator":59,"numerator":200},"profile_key":"native","protocol_description":"Current AA standalone protocol. Status verified in aa-method-current.md section gpqa-diamond","protocol_id":"ac37418879bd94f391def2ae","raw_value":0.305050505050505,"reference_revision":"836004cea651c455b8e60d1d49a8fe83f257d85044bfe4ac3a540a41b9b143b0","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gpqa-diamond"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GPQA Diamond","source_record_id":"minicpm-v4-6-1-3b:gpqa-diamond","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.305050505050505},{"benchmark_family":"GDPval-AA","benchmark_id":"b-40bf5f61ecc772822777","benchmark_label":"GDPval-AA","benchmark_version":"2.1","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"agents","direction":"higher_better","feed_id":"public-aa-current","id":"32822274d56779c973cdcf3ccb12841f","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"v2.1","peer_count":82,"percentile":1.23,"percentile_exact":{"denominator":81,"numerator":100},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; gdpval-aa-normalized","protocol_id":"e895f7b24a99582d5a133a40","raw_value":-362.32,"reference_revision":"0d0b82552cb6c1a35a1a91cb22d1acb0910ad9199fe41bd02da481099d1a15c5","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:gdpval-aa-normalized"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:GDPval-AA","source_record_id":"minicpm-v4-6-1-3b:gdpval-aa-normalized","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"分","value":-362.32},{"benchmark_family":"IFBench","benchmark_id":"b-1cc9c6274c843a5e4097","benchmark_label":"IFBench","benchmark_version":"current published task suite","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"language_information","direction":"higher_better","feed_id":"public-aa-current","id":"e359de5ff096ac45c506a2baeb714ed7","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current 294-item protocol","peer_count":17,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; ifbench","protocol_id":"7b8df779a6b066873d2ea56f","raw_value":0.26734693877551,"reference_revision":"23a287960414276c00683577e559186678df4dc2e82a003aabf8c719538da05f","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:ifbench"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:IFBench","source_record_id":"minicpm-v4-6-1-3b:ifbench","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.26734693877551},{"benchmark_family":"MMMU-Pro","benchmark_id":"b-9196c495a27df3b39ee8","benchmark_label":"MMMU-Pro","benchmark_version":"Standard 10","captured_at":"2026-11-08T03:31:55.736302+00:00","component_key":"score","configuration":"MiniCPM-V 4.6 1.3B","dimension":"multimodal_understanding","direction":"higher_better","feed_id":"public-aa-current","id":"7a567c9ea42d069b6cee3f34d322f5a9","included":true,"model_id":"m-0cd6b9e2d04f2c30","original_version_label":"current Standard 10","peer_count":35,"percentile":0.0,"percentile_exact":{"denominator":1,"numerator":0},"profile_key":"native","protocol_description":"Artificial Analysis current published protocol; mmmu-pro","protocol_id":"9cc385f4bedc10cb3e53b89f","raw_value":0.379190751445087,"reference_revision":"e212153fcce2d1d0cf13094ac801b17118e4f5152b49cf7eeac771eeacb1dc2b","research_artifact_sha256":"6480ada9ca28eb859a34722f14b57e70c1470ad28046ba5f7220a8dfa4649c93","response_sha256":"5a6e86dc8ba82e2fa303d4acfaa22a74e26fffc4a2dd32f3078635611ff397dc","result_date":null,"source_cell":null,"source_id":"s-59fad72c3cdd57202638","source_kind":"independent_evaluation","source_locator":{"image_url":null,"record":"minicpm-v4-6-1-3b:mmmu-pro"},"source_name":"Artificial Analysis","source_protocol_key":"artificial-analysis:current:MMMU-Pro","source_record_id":"minicpm-v4-6-1-3b:mmmu-pro","source_url":"https://artificialanalysis.ai/leaderboards/models","unit":"比例","value":0.379190751445087}],"release_date":"2026-05-11","release_status":"published","release_verified":true,"result_updated_at":null,"score":1.45,"score_exact":{"denominator":19116,"numerator":27625},"source_id":"published/aa:minicpm-v-4-6-1-3b"}],"schema":"model-composite-public-v3","sources":[{"id":"s-59fad72c3cdd57202638","kind":"independent_evaluation","name":"Artificial Analysis","priority":1,"url":"https://artificialanalysis.ai/leaderboards/models"},{"id":"s-094d1db307438c4fd53b","kind":"independent_evaluation","name":"Artificial Analysis","priority":1,"url":"https://artificialanalysis.ai/evaluations/aa-briefcase"},{"id":"s-0a3b665f62276aebfb49","kind":"independent_evaluation","name":"Scale AI","priority":1,"url":"https://labs.scale.com/leaderboard/hle-diamond"},{"id":"s-fbaa49a206de7fcc239b","kind":"independent_evaluation","name":"Scale AI","priority":1,"url":"https://labs.scale.com/leaderboard/hss"},{"id":"s-6e504849b92015a53acf","kind":"independent_evaluation","name":"Scale AI","priority":1,"url":"https://labs.scale.com/leaderboard/mcp_atlas"},{"id":"s-1643719e2a33872a6932","kind":"independent_evaluation","name":"Scale AI","priority":1,"url":"https://labs.scale.com/leaderboard/multichallenge?tab=large"},{"id":"s-e78acdd5f50b212b8328","kind":"independent_evaluation","name":"Epoch AI","priority":1,"url":"https://epoch.ai/data/benchmarks.csv"},{"id":"s-b7fe1fad59fee5842bed","kind":"official_release_comparison","name":"www.anthropic.com","priority":2,"url":"https://www.anthropic.com/claude-opus-5-5"},{"id":"s-c59f3bf1f06fd2e80e5e","kind":"official_release_comparison","name":"www.anthropic.com","priority":2,"url":"https://www.anthropic.com/claude-sonnet-5-5"},{"id":"s-829fcf3ef09cdd6c033e","kind":"official_release_comparison","name":"www.anthropic.com","priority":2,"url":"https://www.anthropic.com/claude-haiku-5-5"},{"id":"s-1c3420336981cb6ce894","kind":"official_release_comparison","name":"www.anthropic.com","priority":2,"url":"https://www.anthropic.com/claude-fable-and-mythos-5-1"},{"id":"s-56f82fed862f1e215c84","kind":"official_release_comparison","name":"openai.com","priority":2,"url":"https://openai.com/index/gpt-6-astra/"},{"id":"s-fb04d3e999c5866880bb","kind":"official_release_comparison","name":"openai.com","priority":2,"url":"https://openai.com/index/gpt-5-6/"},{"id":"s-ac219f0d5e0822daa3e4","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"id":"s-b72308abea4f68de1b3f","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/moonshotai/Kimi-K3"},{"id":"s-ec2bc92acb8dcb2d7ded","kind":"official_release_comparison","name":"qwen.ai","priority":2,"url":"https://qwen.ai/blog?id=qwen3.8"},{"id":"s-e0772012b51061bd757f","kind":"official_release_comparison","name":"qwen.ai","priority":2,"url":"https://qwen.ai/blog?id=qwen3.7"},{"id":"s-b389e587ad922a7e3323","kind":"official_release_comparison","name":"qwen.ai","priority":2,"url":"https://qwen.ai/blog?id=qwen3.7-plus"},{"id":"s-cf62ad1f26e4ee4bb415","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"id":"s-4f49e51ca439fb40ae5c","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"id":"s-94bac3ae47353f5bf78d","kind":"official_release_comparison","name":"deepmind.google","priority":2,"url":"https://deepmind.google/models/model-cards/gemini-3-6-flash/"},{"id":"s-38e2a40ddff67e804e8c","kind":"official_release_comparison","name":"openai.com","priority":2,"url":"https://openai.com/index/introducing-gpt-6-sol-and-luna/"},{"id":"s-095595bcbc77f3c90537","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/LiquidAI/LFM2.5-2.6B"},{"id":"s-6bc244425c066f47875a","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/LiquidAI/LFM2.5-8B-A1B"},{"id":"s-0daea7ae978f1cf387ad","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/upstage/Solar-Open2-250B"},{"id":"s-2d4bd7b6b48d97ab2124","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/openbmb/MiniCPM5-2B"},{"id":"s-ccebc4eeb29cd6842d0c","kind":"official_release_comparison","name":"huggingface.co","priority":2,"url":"https://huggingface.co/tencent/Hy4-preview"},{"id":"s-2dc8e77e0b1e357b6075","kind":"independent_evaluation","name":"Vals AI","priority":1,"url":"https://www.vals.ai/benchmarks/code-migration-v1-1"},{"id":"s-549f084101bc19fe9031","kind":"independent_evaluation","name":"Vals AI","priority":1,"url":"https://www.vals.ai/benchmarks/terminal-bench-4"},{"id":"s-eb560d0cf6499ef1c9f9","kind":"independent_evaluation","name":"Vals AI","priority":1,"url":"https://www.vals.ai/benchmarks/terminal-bench-science"},{"id":"s-ace8373a102d94a857d5","kind":"official_release_comparison","name":"blog.google","priority":2,"url":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/"},{"id":"s-tencent-hy4-appendix-20261106","kind":"official_release","name":"腾讯 Hy4 官方发布比较表","priority":2,"url":"https://huggingface.co/tencent/Hy4-preview"}],"specialized":[],"specialized_recent_count":0,"window":{"end":"2026-11-08","start":"2026-05-08","timezone":"Asia/Shanghai"}}